Building the Silicon Brain - with Drew Houston of Dropbox

CEOs of publicly traded companies are often in the news talking about their new AI initiatives, but few of them have built anything with it. Drew Houston https://x.com/drewhouston?lang=en from Dropbox is different; he has spent over 400 hours coding with LLMs in the last year and is now refocusing his 2,500+ employees around this new way of working, 17 years after founding the company. Timestamps 00:00 Introductions 00:43 Drew's AI journey 04:14 Revalidating expectations of AI 08:23 Simulation in self-driving vs. knowledge work 12:14 Drew's AI Engineering setup 15:24 RAG vs. long context in AI models 18:06 From "FileGPT" to Dropbox AI 23:20 Is storage solved?26:30 Products vs Features 30:48 Building trust for data access 33:42 Dropbox Dash and universal search 38:05 The evolution of Dropbox 42:39 Building a "silicon brain" for knowledge work 48:45 Open source AI and its impact 51:30 "Rent, Don't Buy" for AI 54:50 Staying relevant 58:57 Founder Mode 01:03:10 Advice for founders navigating AI 01:07:36 Building and managing teams in a growing company Transcript Alessio [00:00:00]: Hey everyone, welcome to the Latent Space podcast. This is Alessio, partner and CTO at Decibel Partners, and there's no Swyx today, but I'm joined by Drew Houston of Dropbox. Welcome, Drew. Drew [00:00:14]: Thanks for having me. Alessio [00:00:15]: So we're not going to talk about the Dropbox story. We're not going to talk about the Chinatown bus and the flash drive and all that. I think you've talked enough about it. Where I want to start is you as an AI engineer. So as you know, most of our audience is engineering folks, kind of like technology leaders. You obviously run Dropbox, which is a huge company, but you also do a lot of coding. I think that's how you spend almost 400 hours, just like coding. So let's start there. What was the first interaction you had with an LLM API and when did the journey start for you? Drew [00:00:43]: Yeah. Well, I think probably all AI engineers or whatever you call an AI engineer, those people started out as engineers before that. So engineering is my first love. I mean, I grew up as a little kid. I was that kid. My first line of code was at five years old. I just really loved, I wanted to make computer games, like this whole path. That also led me into startups and eventually starting Dropbox. And then with AI specifically, I studied computer science, I got my, I did my undergrad, but I didn't do like grad level computer science. I didn't, I sort of got distracted by all the startup things, so I didn't do grad level work. But about several years ago, I made a couple of things. So one is I sort of, I knew I wanted to go from being an engineer to a founder. And then, but sort of the becoming a CEO part was sort of backed into the job. And so a couple of realizations. One is that, I mean, there's a lot of like repetitive and like manual work you have to do as an executive that is actually lends itself pretty well to automation, both for like my own convenience. And then out of interest in learning, I guess what we call like classical machine learning these days, I started really trying to wrap my head around understanding machine learning and informational retrieval more, more formally. So I'd say maybe 2016, 2017 started me writing these more successively, more elaborate scripts to like understand basic like classifiers and regression and, and again, like basic information retrieval and NLP back in those days. And there's sort of like two things that came out of that. One is techniques are super powerful. And even just like studying like old school machine learning was a pretty big inversion of the way I had learned engineering, right? You know, I started programming when everyone starts programming and you're, you're sort of the human, you're giving an algorithm to the, and spelling out to the computer how it should run it. And then machine learning, here's machine learning where it's like actually flip that, like give it sort of the answer you want and it'll figure out the algorithm, which was pretty mind bending. And it was both like pretty powerful when I would write tools, like figure out like time audits or like, where's my time going? Is this meeting a one-on-one or is it a recruiting thing or is it a product strategy thing? I started out doing that manually with my assistant, but then found that this was like a very like automatable task. And so, which also had the side effect of teaching me a lot about machine learning. But then there was this big problem, like anytime you, it was very good at like tabular structured data, but like anytime it hit, you know, the usual malformed English that humans speak, it would just like fall over. I had to kind of abandon a lot of the things that I wanted to build because like there's no way to like parse text. Like maybe it would sort of identify the part of speech in a sentence or something. But then fast forward to the LLM, I mean actually I started trying some of like this, what we would call like very small LLMs before kind of the GPT class models. And it was like super hard to get those things working. So like these 500 parameter models would just be like hallucinating and repeating and you know. So actually I'd kind of like written it off a little bit. But then the chat GPT launch and GPT-3 for sure. And then once people figured out like prompting and instruction tuning, this was sort of like November-ish 2022 like everybody else sort of that the chat GPT launch being the starting gun for the whole AI era of computing and then having API access to three and then early access to GPT-4. I was like, oh man, it's happening. And so I was literally on my honeymoon and we're like on a beach in Thailand and I'm like coding these like AI tools to automate like writing or to assist with writing and all these different use cases. Alessio [00:04:14]: You're like, I'm never going back to work. I'm going to automate all of it before I get back. Drew [00:04:17]: And I was just, you know, ever since then, I mean, I've always been like coding like prototypes and just stuff to make my life more convenient, but like escalated a lot after 22. And yeah, I spent, I checked, I think it was probably like over 400 hours this year so far coding because I had my paternity leave where I was able to work on some special projects. But yeah, it's a super important part of like my whole learning journey is like being really hands-on with these things. And I mean, it's probably not a typical recipe, but I really love to get down to the metal as far as how this stuff works. Alessio [00:04:47]: Yeah. So Swyx and I were with Sam Altman in October 22. We were like at a hack day at OpenAI and that's why we started this podcast eventually. But you did an interview with Sam like seven years ago and he asked you what's the biggest opportunity in startups and you were like machine learning and AI and you were almost like too early, right? It's like maybe seven years ago, the models weren't quite there. How should people think about revalidating like expectations of this technology? You know, I think even today people will tell you, oh, models are not really good at X because they were not good 12 months ago, but they're good today. Drew [00:05:19]: What's your project? Heuristics for thinking about that or how is, yeah, I think the way I look at it now is pretty, has evolved a lot since when I started. I mean, I think everybody intuitively starts with like, all right, let's try to predict the future or imagine like what's this great end state we're going to get to. And the tricky thing is like often those prognostications are right, but they're right in terms of direction, but not when. For example, you know, even in the early days of the internet, 90s when things were even like tech space and you know, even before like the browser or things like that, people were like, oh man, you're going to have, you know, you're going to be able to order food, get like a Snickers delivered to your house, you're going to be able to watch any movie ever created. And they were right. But they were like, you know, it took 20 years for that to actually happen. And before you got to DoorDash, you had to get, you started with like Webvan and Cosmo and before you get to Spotify, you had to do like Napster and Kazaa and LimeWire and like a bunch of like broken Britney Spears MP3s and malware. So I think the big lesson is being early is the same as being wrong. Being late is the same as being wrong. So really how do you calibrate timing? And then I think with AI, it's the same thing that people are like, oh, it's going to completely upend society and all these positive and negative ways. I think that's like most of those things are going to come true. The question is like, when is that going to happen? And then with AI specifically, I think there's also, in addition to sort of the general tech category or like jumping too fast to the future, I think that AI is particularly susceptible to that. And you look at self-driving, right? This idea of like, oh my God, you can have a self-driving car captured everybody's imaginations 10, 12 years ago. And you know, people are like, oh man, in two years, there's not going to be another year. There's not going to be a human driver on the road to be seen. It didn't work out that way, right? We're still 10, 12 years later where we're in a world where you can sort of sometimes get a Waymo in like one city on earth. Exciting, but just took a lot longer than people think. And the reason is there's a lot of engineering challenges, but then there's a lot of other like societal time constants that are hard to compress. So one thing I think you can learn from things like self-driving is they have these levels of autonomy that's a useful kind of framework in driving or these like maturity levels. People sort of skip to like level five, full autonomy, or we're going to have like an autonomous knowledge worker that's just going to take, that's going to, and then we won't need humans anymore kind of projection that that's going to take a long time. But then when you think about level one or level two, like these little assistive experiences, you know, we're seeing a lot of traction with those. So what you see really working is the level one autonomy in the AI world would be like the tab auto-complete and co-pilot, right? And then, you know, maybe a little higher is like the chatbot type interface. Obviously you want to get to the highest level you can to build a good product, but the reliability just isn't, and the capability just isn't there in the early innings. And so, and then you think of other level one, level two type things, like Google Maps probably did more for self-driving than in literal self-driving, like a billion people have like the ability to have like maps and navigation just like taken care of for you autonomously. So I think the timing and maturity are really important factors to include. Alessio [00:08:23]: The thing with self-driving, maybe one of the big breakthroughs was like simulation. So it's like, okay, instead of driving, we can simulate these environments. It's really hard to do when knowledge work, you know, how do you simulate like a product review? How do you simulate these things? I'm curious if you've done any experiments. I know some companies have started to build kind of like a virtual personas that you can like bounce ideas off of. Drew [00:08:42]: I mean, fortunately in a company you generate lots of, you know, actual human training data all the time. And then I also just like start with myself, like, all right, I can, you know, it's pretty tricky even within your company to be like, all right, let's open all this up as quote training data. But, you know, I can start with my own emails or my own calendar or own stuff without running into the same kind of like privacy or other concerns. So I often like start with my own stuff. And so that is like a one level of bootstrapping, but actually four or five years ago during COVID, we decided, you know, a lot of companies were thinking about how do we go back to work? And so we decided to really lean into remote and distributed work because I thought, you know, this is going to be the biggest change to the way we work in our lifetimes. And COVID kind of ripped up a bunch of things, but I think everybody was sort of pleasantly surprised how with a lot of knowledge work, you could just keep going. And actually you were sort of fine. Work was decoupled from your physical environment, from being in a physical place, which meant that things people had dreamed about since the fifties or sixties, like telework, like you actually could work from anywhere. And that was now possible. So we decided to really lean into that because we debated, should we sort of hit the fast forward button or should we hit the rewind button and go back to 2019? And obviously that's been playing out over the last few years. And we decided to basically turn, we went like 90% remote. We still, the in-person part's really important. We can kind of come back to our working model, but we're like, yeah, this is, everybody is going to be in some kind of like distributed or hybrid state. So like instead of like running away from this, like let's do a full send, let's really go into it. Let's live in the future. A few years before our customers, let's like turn Dropbox into a lab for distributed work. And we do that like quite literally, both of the working model and then increasingly with our products. And then absolutely, like we have products like Dropbox Dash, which is our universal search product. That was like very elevated in priority for me after COVID because like now you have, we're putting a lot more stress on the system and on our screens, it's a lot more chaotic and overwhelming. And so even just like getting the right information, the right person at the right time is a big fundamental challenge in knowledge work and these, in the distributed world, like big problem today is still getting, you know, has been getting bigger. And then for a lot of these other workflows, yeah, there's, we can both get a lot of natural like training data from just our own like strategy docs and processes. There's obviously a lot you can do with synthetic data and you know, actually like LMs are pretty good at being like imitating generic knowledge workers. So it's, it's kind of funny that way, but yeah, the way I look at it is like really turn Dropbox into a lab for distributed work. You think about things like what are the big problems we're going to have? It's just the complexity on our screens just keeps growing and the whole environment gets kind of more out of sync with what makes us like cognitively productive and engaged. And then even something like Dash was initially seeded, I made a little personal search engine because I was just like personally frustrated with not being able to find my stuff. And along that whole learning journey with AI, like the vector search or semantic search, things like that had just been the tooling for that. The open source stuff had finally gotten to a place where it was a pretty good developer experience. And so, you know, in a few days I had sort of a hello world type search engine and I'm like, oh my God, like this completely works. You don't even have to get the keywords right. The relevance and ranking is super good. We even like untuned. So I guess that's to say like I've been surprised by if you choose like the right algorithm and the right approach, you can actually get like super good results without having like a ton of data. And even with LLMs, you can apply all these other techniques to give them, kind of bootstrap kind of like task maturity pretty quickly. Alessio [00:12:14]: Before we jump into Dash, let's talk about the Drew Haas and AI engineering stuff. So IDE, let's break that down. What IDE do you use? Do you use Cursor, VS Code, do you use any coding assistant, like WeChat, is it just autocomplete? Drew [00:12:28]: Yeah, yeah. Both. So I use VS Code as like my daily driver, although I'm like super excited about things like Cursor or the AI agents. I have my own like stack underneath that. I mean, some off the shelf parts, some pretty custom. So I use the continue.dev just like AI chat UI basically as just the UI layer, but I also proxy the request. I proxy the request to my own backend, which is sort of like a router. You can use any backend. I mean, Sonnet 3.5 is probably the best all around. But then these things are like pretty limited if you don't give them the right context. And so part of what the proxy does is like there's a separate thing where I can say like include all these files by default with the request. And then it becomes a lot easier and like without like cutting and pasting. And I'm building mostly like prototype toy apps, so it's like a front end React thing and a Python backend thing. And so it can do these like end to end diffs basically. And then I also like love being able to host everything locally or do it offline. So I have my own, when I'm on a plane or something or where like you don't have access or the internet's not reliable, I actually bring a gaming laptop on the plane with me. It's like a little like blue briefcase looking thing. And then I like literally hook up a GPU like into one of the outlets. And then I have, I can do like transcription, I can do like autocomplete, like I have an 8 billion, like Llama will run fine. Alessio [00:13:44]: And you're using like a Llama to run the model? Drew [00:13:47]: No, I use, I have my own like LLM inference stack. I mean, it uses the backend somewhat interchangeable. So everything from like XLlama to VLLM or SGLang, there's a bunch of these different backends you can use. And then I started like working on stuff before all this tooling was like really available. So you know, over the last several years, I've built like my own like whole crazy environment and like in stack here. So I'm a little nuts about it. Alessio [00:14:12]: Yeah. What's the state of the art for, I guess not state of the art, but like when it comes to like frameworks and things like that, do you like using them? I think maybe a lot of people say, hey, things change so quickly, they're like trying to abstract things. Yeah. Drew [00:14:24]: It's maybe too early today. As much as I do a lot of coding, I have to be pretty surgical with my time. I don't have that much time, which means I have to sort of like scope my innovation to like very specific places or like my time. So for the front end, it'll be like a pretty vanilla stack, like a Next.js, React based thing. And then these are toy apps. So it's like Python, Flask, SQLite, and then all the different, there's a whole other thing on like the backend. Like how do you get, sort of run all these models locally or with a local GPU? The scaffolding on the front end is pretty straightforward, the scaffolding on the backend is pretty straightforward. Then a lot of it is just like the LLM inference and control over like fine grained aspects of how you do generation, caching, things like that. And then there's a lot, like a lot of the work is how do you take, sort of go to an IMAP, like take an email, get a new, or a document or a spreadsheet or any of these kinds of primitives that you work with and then translate them, render them in a format that an LLM can understand. So there's like a lot of work that goes into that too. Yeah. Alessio [00:15:24]: So I built a kind of like email triage system and like I would say 80% of the code is like Google and like pulling emails and then the actual AI part is pretty easy. Drew [00:15:34]: Yeah. And even, same experience. And then I tried to do all these like NLP things and then to my dismay, like a bunch of reg Xs were like, got you like 95% of the way there. So I still leave it running, I just haven't really built like the LLM powered version of it yet. Yeah. Alessio [00:15:51]: So do you have any thoughts on rag versus long context, especially, I mean with Dropbox, you know? Sure. Do you just want to shove things in? Like have you seen that be a lot better? Drew [00:15:59]: Well, they kind of have different strengths and weaknesses, so you need both for different use cases. I mean, it's been awesome in the last 12 months, like now you have these like long context models that can actually do a lot. You can put a book in, you know, Sonnet's context and then now with the later versions of LLAMA, you can have 128k context. So that's sort of the new normal, which is awesome and that, that wasn't even the case a year ago. That said, models don't always use, and certainly like local models don't use the full context well fully yet, and actually if you provide too much irrelevant context, the quality degrades a lot. And so I say in the open source world, like we're still just getting to the cusp of like the full context is usable. And then of course, like when you're something like Dropbox Dash, like it's basically building this whole like brain that's like read everything your company's ever written. And so that's not going to fit into your context window, so you need rag just as a practical reality. And even for a lot of similar reasons, you need like RAM and hard disk in conventional computer architecture. And I think these things will keep like horse trading, like maybe if, you know, a million or 10 million is the new, tokens is the new context length, maybe that shifts. Maybe the bigger picture is like, it's super exciting to talk about the LLM and like that piece of the puzzle, but there's this whole other scaffolding of more conventional like retrieval or conventional machine learning, especially because you have to scale up products to like millions of people you do in your toy app is not going to scale to that from a cost or latency or performance standpoint. So I think you really need these like hybrid architectures that where you have very like purpose fit tools, or you're probably not using Sonnet 3.5 for all of your normal product use cases. You're going to use like a fine tuned 8 billion model or sort of the minimum model that gets you the right output. And then a smaller model also is like a lot more cost and latency versus like much better characteristics on that front. Alessio [00:17:48]: Yeah. Let's jump into the Dropbox AI story. So sure. Your initial prototype was Files GPT. How did it start? And then how did you communicate that internally? You know, I know you have a pretty strong like mammal culture. One where you're like, okay, Hey, we got to really take this seriously. Drew [00:18:06]: Yeah. Well, on the latter, it was, so how do we say like how we took Dropbox, how AI seriously as a company started kind of around that time, that honeymoon time, unfortunately. In January, I wrote this like memo to the company, like around basically like how we need to play offense in 23. And that most of the time the kind of concrete is set and like the winners are the winners and things are kind of frozen. But then with these new eras of computing, like the PC or the internet or the phone or the concrete on freezes and you can sort of build, do things differently and have a new set of winners. It's sort of like a new season starts as a result of a lot of that sort of personal hacking and just like thinking about this. I'm like, yeah, this is an inflection point in the industry. Like we really need to change how we think about our strategy. And then becoming an AI first company was probably the headline thing that we did. And then, and then that got, and then calling on everybody in the company to really think about in your world, how is AI going to reshape your workflows or what sort of the AI native way of thinking about your job. File GPT, which is sort of this Dropbox AI kind of initial concept that actually came from our engineering team as, you know, as we like called on everybody, like really think about what we should be doing that's new or different. So it was kind of organic and bottoms up like a bunch of engineers just kind of hacked that together. And then that materialized as basically when you preview a file on Dropbox, you can have kind of the most straightforward possible integration of AI, which is a good thing. Like basically you have a long PDF, you want to be able to ask questions of it. So like a pretty basic implementation of RAG and being able to do that when you preview a file on Dropbox. So that was the origin of that, that was like back in 2023 when we released just like the starting engines had just, you know, gotten going. Alessio [00:19:53]: It's funny where you're basically like these files that people have, they really don't want them in a way, you know, like you're storing all these files and like you actually don't want to interact with them. You want a layer on top of it. And that's kind of what also takes you to Dash eventually, which is like, Hey, you actually don't really care where the file is. You just want to be the place that aggregates it. How do you think about what people will know about files? You know, are files the actual file? Are files like the metadata and they're just kind of like a pointer that goes somewhere and you don't really care where it is? Drew [00:20:21]: Yeah. Alessio [00:20:22]: Any thoughts about? Drew [00:20:23]: Totally. Yeah. I mean, there's a lot of potential complexity in that question, right? Is it a, you know, what's the difference between a file and a URL? And you can go into the technicals, it's like pass by value, pass by reference. Okay. What's the format like? All right. So it starts with a primitive. It's not really a flat file. It's like a structured data. You're sort of collaborative. Yeah. That's keeping in sync. Blah, blah, blah. I actually don't start there at all. I just start with like, what do people, like, what do humans, let's work back from like how humans think about this stuff or how they should think about this stuff. Meaning like, I don't think about, Oh, here are my files and here are my links or cloud docs. I'm just sort of like, Oh, here's my stuff. This, this, here's sort of my documents. Here's my media. Here's my projects. Here are the people I'm working with. So it starts from primitives more like those, like how do people, how do humans think about these things? And then, then start from like a more ideal experience. Because if you think about it, we kind of have this situation that will look like particularly medieval in hindsight where, all right, how do you manage your work stuff? Well, on all, you know, on one side of your screen, you have this file browser that literally hasn't changed since the early eighties, right? You could take someone from the original Mac and sit them in front of like a computer and they'd be like, this is it. And that's, it's been 40 years, right? Then on the other side of your screen, you have like Chrome or a browser that has so many tabs open, you can no longer see text or titles. This is the state of the art for how we manage stuff at work. Interestingly, neither of those experiences was purpose-built to be like the home for your work stuff or even anything related to it. And so it's important to remember, we get like stuck in these local maxima pretty often in tech where we're obviously aware that files are not going away, especially in certain domains. So that format really matters and where files are still going to be the tool you use for like if there's something big, right? If you're a big video file, that kind of format in a file makes sense. There's a bunch of industries where it's like construction or architecture or sort of these domain specific areas, you know, media generally, if you're making music or photos or video, that all kind of fits in the big file zone where Dropbox is really strong and that's like what customers love us for. It's also pretty obvious that a lot of stuff that used to be in, you know, Word docs or Excel files, like all that has tilted towards the browser and that tilt is going to continue. So with Dash, we wanted to make something that was really like cloud-native, AI-native and deliberately like not be tied down to the abstractions of the file system. Now on the other hand, it would be like ironic and bad if we then like fractured the experience that you're like, well, if it touches a file, it's a syncing metaphor to this app. And if it's a URL, it's like this completely different interface. So there's a convergence that I think makes sense over time. But you know, but I think you have to start from like, not so much the technology, start from like, what do the humans want? And then like, what's the idealized product experience? And then like, what are the technical underpinnings of that, that can make that good experience? Alessio [00:23:20]: I think it's kind of intuitive that in Dash, you can connect Google Drive, right? Because you think about Dropbox, it's like, well, it's file storage, you really don't want people to store files somewhere, but the reality is that they do. How do you think about the importance of storage and like, do you kind of feel storage is like almost solved, where it's like, hey, you can kind of store these files anywhere, what matters is like access. Drew [00:23:38]: It's a little bit nuanced in that if you're dealing with like large quantities of data, it actually does matter. The implementation matters a lot or like you're dealing with like, you know, 10 gig video files like that, then you sort of inherit all the problems of sync and have to go into a lot of the challenges that we've solved. Switching on a pretty important question, like what is the value we provide? What does Dropbox do? And probably like most people, I would have said like, well, Dropbox syncs your files. And we didn't even really have a mission of the company in the beginning. I'm just like, yeah, I just don't want to carry a thumb driving around and life would be a lot better if our stuff just like lived in the cloud and I just didn't have to think about like, what device is the thing on or what operating, why are these operating systems fighting with each other and incompatible? You know, I just want to abstract all of that away. But then so we thought, even we were like, all right, Dropbox provides storage. But when we talked to our customers, they're like, that's not how we see this at all. Like actually, Dropbox is not just like a hard drive in the cloud. It's like the place where I go to work or it's a place like I started a small business is a place where my dreams come true. Or it's like, yeah, it's not keeping files in sync. It's keeping people in sync. It's keeping my team in sync. And so they're using this kind of language where we're like, wait, okay, yeah, because I don't know, storage probably is a commodity or what we do is a commodity. But then we talked to our customers like, no, we're not buying the storage, we're buying like the ability to access all of our stuff in one place. We're buying the ability to share everything and sort of, in a lot of ways, people are buying the ability to work from anywhere. And Dropbox was kind of, the fact that it was like file syncing was an implementation detail of this higher order need that they had. So I think that's where we start too, which is like, what is the sort of higher order thing, the job the customer is hiring Dropbox to do? Storage in the new world is kind of incidental to that. I mean, it still matters for things like video or those kinds of workflows. The value of Dropbox had never been, we provide you like the cheapest bits in the cloud. But it is a big pivot from Dropbox is the company that syncs your files to now where we're going is Dropbox is the company that kind of helps you organize all your cloud content. I started the company because I kept forgetting my thumb drive. But the question I was really asking was like, why is it so hard to like find my stuff, organize my stuff, share my stuff, keep my stuff safe? You know, I'm always like one washing machine and I would leave like my little thumb drive with all my prior company stuff on in the pocket of my shorts and then almost wash it and destroy it. And so I was like, why do we have to, this is like medieval that we have to think about this. So that same mindset is how I approach where we're going. But I think, and then unfortunately the, we're sort of back to the same problems. Like it's really hard to find my stuff. It's really hard to organize myself. It's hard to share my stuff. It's hard to secure my content at work. Now the problem is the same, the shape of the problem and the shape of the solution is pretty different. You know, instead of a hundred files on your desktop, it's now a hundred tabs in your browser, et cetera. But I think that's the starting point. Alessio [00:26:30]: How has the idea of a product evolved for you? So, you know, famously Steve Jobs started by Dropbox and he's like, you know, this is just a feature. It's not a product. And then you build like a $10 billion feature. How in the age of AI, how do you think about, you know, maybe things that used to be a product are now features because the AI on top of it, it's like the product, like what's your mental model? Do you think about it? Drew [00:26:50]: Yeah. So I don't think there's really like a bright line. I don't know if like I use the word features and products and my mental model that much of how I break it down because it's kind of a, it's a good question. I mean, I don't not think about features, I don't think about products, but it does start from that place of like, all right, we have all these new colors we can paint with and all right, what are these higher order needs that are sort of evergreen, right? So people will always have stuff at work. They're always need to be able to find it or, you know, all the verbs I just mentioned. It's like, okay, how can we make like a better painting and how can we, and then how can we use some of these new colors? And then, yeah, it's like pretty clear that after the large models, the way you find stuff share stuff, it's going to be completely different after COVID, it's going to be completely different. So that's the starting point. But I think it is also important to, you know, you have to do more than just work back from the customer and like what they're trying to do. Like you have to think about, and you know, we've, we've learned a lot of this the hard way sometimes. Okay. You might start with a customer. You might start with a job to be on there. You're like, all right, what's the solution to their problem? Or like, can we build the best product that solves that problem? Right. Like what's the best way to find your stuff in the modern world? Like, well, yeah, right now the status quo for the vast majority of the billion, billion knowledge workers is they have like 10 search boxes at work that each search 10% of your stuff. Like that's clearly broken. Obviously you should just have like one search box. All right. So we can do that. And that also has to be like, I'll come back to defensibility in a second, but like, can we build the right solution that is like meaningfully better from the status quo? Like, yes, clearly. Okay. Then can we like get distribution and growth? Like that's sort of the next thing you learned is as a founder, you start with like, what's the product? What's the product? What's the product? Then you're like, wait, wait, we need distribution and we need a business model. So those are the next kind of two dominoes you have to knock down or sort of needles you have to thread at the same time. So all right, how do we grow? I mean, if Dropbox 1.0 is really this like self-serve viral model that there's a lot of, we sort of took a borrowed from a lot of the consumer internet playbook and like what Facebook and social media were doing and then translated that to sort of the business world. How do you get distribution, especially as a startup? And then a business model, like, all right, storage happened to be something in the beginning happened to be something people were willing to pay for. They recognize that, you know, okay, if I don't buy something like Dropbox, I'm going to have to buy an external hard drive. I'm going to have to buy a thumb drive and I have to pay for something one way or another. People are already paying for things like backup. So we felt good about that. But then the last domino is like defensibility. Okay. So you build this product or you get the business model, but then, you know, what do you do when the incumbents, the next chess move for them is I just like copy, bundle, kill. So they're going to copy your product. They'll bundle it with their platforms and they'll like give it away for free or no added cost. And, you know, we had a lot of, you know, scar tissue from being on the wrong side of that. Now you don't need to solve all four for all four or five variables or whatever at once or you can sort of have, you know, some flexibility. But the more of those gates that you get through, you sort of add a 10 X to your valuation. And so with AI, I think, you know, there's been a lot of focus on the large language model, but it's like large language models are a pretty bad business from a, you know, you sort of take off your tech lens and just sort of business lens. Like there's sort of this weirdly self-commoditizing thing where, you know, models only have value if they're kind of on this like Pareto frontier of size and quality and cost. Being number two, you know, if you're not on that frontier, the second the frontier moves out, which it moves out every week, like your model literally has zero economic value because it's dominated by the new thing. LLMs generate output that can be used to train or improve. So there's weird, peculiar things that are specific to the large language model. And then you have to like be like, all right, where's the value going to accrue in the stack or the value chain? And, you know, certainly at the bottom with Nvidia and the semiconductor companies, and then it's going to be at the top, like the people who have the customer relationship who have the application layer. Those are a few of the like lenses that I look at a question like that through. Alessio [00:30:48]: Do you think AI is making people more careful about sharing the data at all? People are like, oh, data is important, but it's like, whatever, I'm just throwing it out there. Now everybody's like, but are you going to train on my data? And like your data is actually not that good to train on anyway. But like how have you seen, especially customers, like think about what to put in, what to not? Drew [00:31:06]: I mean, everybody should be. Well, everybody is concerned about this and nobody should be concerned about this, right? Because nobody wants their personal companies information to be kind of ground up into little pellets to like sell you ads or train the next foundation model. I think it's like massively top of mind for every one of our customers, like, and me personally, and with my Dropbox hat on, it's like so fundamental. And, you know, we had experience with this too at Dropbox 1.0, the same kind of resistance, like, wait, I'm going to take my stuff on my hard drive and put it on your server somewhere. Are you serious? What could possibly go wrong? And you know, before that, I was like, wait, are you going to sell me, I'm going to put my credit card number into this website? And before that, I was like, hey, I'm going to take all my cash and put it in a bank instead of under my mattress. You know, so there's a long history of like tech and comfort. So in some sense, AI is kind of another round of the same thing, but the issues are real. And then when I think about like defensibility for Dropbox, like that's actually a big advantage that we have is one, our incentives are very aligned with our customers, right? We only get, we only make money if you pay us and you only pay us if we do a good job. So we don't have any like side hustle, you know, we're not training the next foundation model. You know, we're not trying to sell you ads. Actually we're not even trying to lock you into an ecosystem, like the whole point of Dropbox is it works, you know, everywhere. Because I think one of the big questions we've circling around is sort of like, in the world of AI, where should our lane be? Like every startup has to ask, or in every big company has to ask, like, where can we really win? But to me, it was like a lot of the like trust advantages, platform agnostic, having like a very clean business model, not having these other incentives. And then we also are like super transparent. We were transparent early on. We're like, all right, we're going to establish these AI principles, very table stakes stuff of like, here's transparency. We want to give people control. We want to cover privacy, safety, bias, like fairness, all these things. And we put that out up front to put some sort of explicit guardrails out where like, hey, we're, you know, because everybody wants like a trusted partner as they sort of go into the wild world of AI. And then, you know, you also see people cutting corners and, you know, or just there's a lot of uncertainty or, you know, moving the pieces around after the fact, which no one feels good about. Alessio [00:33:14]: I mean, I would say the last 10, 15 years, the race was kind of being the system of record, being the storage provider. I think today it's almost like, hey, if I can use Dash to like access my Google Drive file, why would I pay Google for like their AI feature? So like vice versa, you know, if I can connect my Dropbook storage to this other AI assistant, how do you kind of think about that, about, you know, not being able to capture all the value and how open people will stay? I think today things are still pretty open, but I'm curious if you think things will get more closed or like more open later. Drew [00:33:42]: Yeah. Well, I think you have to get the value exchange right. And I think you have to be like a trustworthy partner or like no one's going to partner with you if they think you're going to eat their lunch, right? Or if you're going to disintermediate them and like all the companies are quite sophisticated with how they think about that. So we try to, like, we know that's going to be the reality. So we're actually not trying to eat anyone's like Google Drive's lunch or anything. Actually we'll like integrate with Google Drive, we'll integrate with OneDrive, really any of the content platforms, even if they compete with file syncing. So that's actually a big strategic shift. We're not really reliant on being like the store of record and there are pros and cons to this decision. But if you think about it, we're basically like providing all these apps more engagement. We're like helping users do what they're really trying to do, which is to get, you know, that Google Doc or whatever. And we're not trying to be like, oh, by the way, use this other thing. This is all part of our like brand reputation. It's like, no, we give people freedom to use whatever tools or operating system they want. We're not taking anything away from our partners. We're actually like making it, making their thing more useful or routing people to those things. I mean, on the margin, then we have something like, well, okay, to the extent you do rag and summarize things, maybe that doesn't generate a click. Okay. You know, we also know there's like infinity investment going into like the work agents. So we're not really building like a co-pilot or Gemini competitor. Not because we don't like those. We don't find that thing like captivating. Yeah, of course. But just like, you know, you learn after some time in this business that like, yeah, there's some places that are just going to be such kind of red oceans or just like super big battlefields. Everybody's kind of trying to solve the same problem and they just start duplicating all each other effort. And then meanwhile, you know, I think the concern would be is like, well, there's all these other problems that aren't being properly addressed by AI. And I was concerned that like, yeah, and everybody's like fixated on the agent or the chatbot interface, but forgetting that like, hey guys, like we have the opportunity to like really fix search or build a self-organizing Dropbox or environment or there's all these other things that can be a compliment. Because we don't really want our customers to be thinking like, well, do I use Dash or do I use co-pilot? And frankly, none of them do. In a lot of ways, actually, some of the things that we do on the security front with Dash for Business are a good compliment to co-pilot. Because as part of Dash for Business, we actually give admins, IT, like universal visibility and control over all the different, what's being shared in your company across all these different platforms. And as a precondition to installing something like co-pilot or Dash or Glean or any of these other things, right? You know, IT wants to know like, hey, before we like turn all the lights in here, like let's do a little cleaning first before we let everybody in. And there just haven't been good tools to do that. And post AI, you would do it completely differently. And so that's like a big, that's a cornerstone of what we do and what sets us apart from these tools. And actually, in a lot of cases, we will help those tools be adopted because we actually help them do it safely. Yeah. Alessio [00:36:27]: How do you think about building for AI versus people? It's like when you mentioned cleaning up is because maybe before you were like, well, humans can have some common sense when they look at data on what to pick versus models are just kind of like ingesting. Do you think about building products differently, knowing that a lot of the data will actually be consumed by LLMs and like agents and whatnot versus like just people? Drew [00:36:46]: I think it'll always be, I aim a little bit more for like, you know, level three, level four kind of automation, because even if the LLM is like capable of completely autonomously organizing your environment, it probably would do a reasonable job. But like, I think you build bad UI when the sort of user has to fit itself to the computer versus something that you're, you know, it's like an instrument you're playing or something where you have some kind of good partnership. And you know, and on the other side, you don't have to do all this like manual effort. And so like the command line was sort of subsumed by like, you know, graphical UI. We'll keep toggling back and forth. Maybe chat will be, chat will be an increasing, especially when you bring in voice, like will be an increasing part of the puzzle. But I don't think we're going to go back to like a million command lines either. And then as far as like the sort of plumbing of like, well, is this going to be consumed by an LLM or a human? Like fortunately, like you don't really have to design it that differently. I mean, you have to make sure everything's legible to the LLM, but it's like quite tolerant of, you know, malformed everything. And actually the more, the easier it makes something to read for a human, the easier it is for an LLM to read to some extent as well. But we really think about what's that kind of right, how do we build that right, like human machine interface where you're still in control and driving, but then it's super easy to translate your intent into like the, you know, however you want your folder, setting your environment set up or like your preferences. Alessio [00:38:05]: What's the most underrated thing about Dropbox that maybe people don't appreciate? Drew [00:38:09]: Well, I think this is just such a natural evolution for us. It's pretty true. Like when people think about the world of AI, file syncing is not like the next thing you would auto complete mentally. And I think we also did like our first thing so well that there were a lot of benefits to that. But I think there also are like, we hit it so hard with our first product that it was like pretty tough to come up with a sequel. And we had a bit of a sophomore slump and you know, I think actually a lot of kids do use Dropbox through in high school or things like that, but you know, they're not, they're using, they're a lot more in the browser and then their file system, right. And we know all this, but still like we're super well positioned to like help a new generation of people with these fundamental problems and these like that affect, you know, a billion knowledge workers around just finding, organizing, sharing your stuff and keeping it safe. And there's, there's a ton of unsolved problems in those four verbs. We've talked about search a little bit, but just even think about like a whole new generation of people like growing up without the ability to like organize their things and yeah, search is great. And if you just have like a giant infinite pile of stuff, then search does make that more manageable. But you know, you do lose some things that were pretty helpful in prior decades, right? So even just the idea of persistence, stuff still being there when you come back, like when I go to sleep and wake up, my physical papers are still on my desk. When I reboot my computer, the files are still on my hard drive. But then when in my browser, like if my operating system updates the wrong way and closes the browser or if I just more commonly just declared tab bankruptcy, it's like your whole workspace just clears itself out and starts from zero. And you're like, on what planet is this a good idea? There's no like concept of like, oh, here's the stuff I was working on. Yeah, let me get back to it. And so that's like a big motivation for things like Dash. Huge problems with sharing, right? If I'm remodeling my house or if I'm getting ready for a board meeting, you know, what do I do if I have a Google doc and an air table and a 10 gig 4k video? There's no collection that holds mixed format things. And so it's another kind of hidden problem, hidden in plain sight, like he's missing primitives. Files have folders, songs have playlists, links have, you know, there's no, somehow we miss that. And so we're building that with stacks in Dash where it's like a mixed format, smart collection that you can then, you know, just share whatever you need internally, externally and have it be like a really well designed experience and platform agnostic and not tying you to any one ecosystem. We're super excited about that. You know, we talked a little bit about security in the modern world, like IT signs all these compliance documents, but in reality has no way of knowing where anything is or what's being shared. It's actually better for them to not know about it than to know about it and not be able to do anything about it. And when we talked to customers, we found that there were like literally people in IT whose jobs it is to like manually go through, log into each, like log into office, log into workspace, log into each tool and like go comb through one by one the links that people have shared and like unshares. There's like an unshare guy in all these companies and that that job is probably about as fun as it sounds like, my God. So there's, you know, fortunately, I guess what makes technology a good business is for every problem it solves, it like creates a new one, so there's always like a sequel that you need. And so, you know, I think the happy version of our Act 2 is kind of similar to Netflix. I look at a lot of these companies that really had multiple acts and Netflix had the vision to be streaming from the beginning, but broadband and everything wasn't ready for it. So they started by mailing you DVDs, but then went to streaming and then, but the value probably the whole time was just like, let me press play on something I want to see. And they did a really good job about bringing people along from the DVD mailing off. You would think like, oh, the DVD mailing piece is like this burning platform or it's like legacy, you know, ankle weight. And they did have some false starts in that transition. But when you really think about it, they were able to take that DVD mailing audience, move, like migrate them to streaming and actually bootstrap a, you know, take their season one people and bootstrap a victory in season two, because they already had, you know, they weren't starting from scratch. And like both of those worlds were like super easy to sort of forget and be like, oh, it's all kind of destiny. But like, no, that was like an incredibly competitive environment. And Netflix did a great job of like activating their Act 1 advantages and winning in Act 2 because of it. So I don't think people see Dropbox that way. I think people are sort of thinking about us just in terms of our Act 1 and they're like, yeah, Dropbox is fine. I used to use it 10 years ago. But like, what have they done for me lately? And I don't blame them. So fortunately, we have like better and better answers to that question every year. Alessio [00:42:39]: And you call it like the silicon brain. So you see like Dash and Stacks being like the silicon brain interface, basically for Drew [00:42:46]: people. I mean, that's part of it. Yeah. And writ large, I mean, I think what's so exciting about AI and everybody's got their own kind of take on it, but if you like really zoom out civilizationally and like what allows humans to make progress and, you know, what sort of is above the fold in terms of what's really mattered. I certainly want to, I mean, there are a lot of points, but some that come to mind like you think about things like the industrial revolution, like before that, like mechanical energy, like the only way you could get it was like by your own hands, maybe an animal, maybe some like clever sort of machines or machines made of like wood or something. But you were quite like energy limited. And then suddenly, you know, the industrial revolution, things like electricity, it suddenly is like, all right, mechanical energy is now available on demand as a very fungible kind of, and then suddenly we consume a lot more of it. And then the standard of living goes way, way, way, way up. That's been pretty limited to the physical realm. And then I believe that the large models, that's really the first time we can kind of bottle up cognitive energy and offloaded, you know, if we started by offloading a lot of our mechanical or physical busy work to machines that freed us up to make a lot of progress in other areas. But then with AI and computing, we're like, now we can offload a lot more of our cognitive busy work to machines. And then we can create a lot more of it. Price of it goes way down. Importantly, like, it's not like humans never did anything physical again. It's sort of like, no, but we're more leveraged. We can move a lot more earth with a bulldozer than a shovel. And so that's like what is at the most fundamental level, what's so exciting to me about AI. And so what's the silicon brain? It's like, well, we have our human brains and then we're going to have this other like half of our brain that's sort of coming online, like our silicon brain. And it's not like one or the other. They complement each other. They have very complimentary strengths and weaknesses. And that's, that's a good thing. There's also this weird tangent we've gone on as a species to like where knowledge work, knowledge workers have this like epidemic of, of burnout, great resignation, quiet quitting. And there's a lot going on there. But I think that's one of the biggest problems we have is that be like, people deserve like meaningful work and, you know, can't solve all of it. But like, and at least in knowledge work, there's a lot of own goals, you know, enforced errors that we're doing where it's like, you know, on one side with brain science, like we know what makes us like productive and fortunately it's also what makes us engaged. It's like when we can focus or when we're some kind of flow state, but then we go to work and then increasingly going to work is like going to a screen and you're like, if you wanted to design an environment that made it impossible to ever get into a flow state or ever be able to focus, like what we have is that. And that was the thing that just like seven, eight years ago just blew my mind. I'm just like, I cannot understand why like knowledge work is so jacked up on this adventure. It's like, we, we put ourselves in like the most cognitively polluted environment possible and we put so much more stress on the system when we're working remotely and things like that. And you know, all of these problems are just like going in the wrong direction. And I just, I just couldn't understand why this was like a problem that wasn't fixing itself. And I'm like, maybe there's something Dropbox can do with this and you know, things like Dash are the first step. But then, well, so like what, well, I mean, now like, well, why are humans in this like polluted state? It's like, well, we're just, all of the tools we have today, like this generation of tools just passes on all of the weight, the burden to the human, right? So it's like, here's a bajillion, you know, 80,000 unread emails, cool. Here's 25 unread Slack channels. Here's, we all get started like, it's like jittery like thinking about it. And then you look at that, you're like, wait, I'm looking at my phone, it says like 80,000 unread things. There's like no question, product question for which this is the right answer. Fortunately, that's why things like our silicon brain are pretty helpful because like they can serve as like an attention filter where it's like, actually, computers have no problem reading a million things. Humans can't do that, but computers can. And to some extent, this was already happening with computer, you know, Excel is an aversion of your silicon brain or, you know, you could draw the line arbitrarily. But with larger models, like now so many of these little subtasks and tasks we do at work can be like fully automated. And I think, you know, I think it's like an important metaphor to me because it mirrors a lot of what we saw with computing, computer architecture generally. It's like we started out with the CPU, very general purpose, then GPU came along much better at these like parallel computations. We talk a lot about like human versus machine being like substituting, it's like CPU, GPU, it's not like one is categorically better than the other, they're complements. Like if you have something really parallel, use a GPU, if not, use a CPU. The whole relationship, that symbiosis between CPU and GPU has obviously evolved a lot since, you know, playing Quake 2 or something. But right now we have like the human CPU doing a lot of, you know, silicon CPU tasks. And so you really have to like redesign the work thoughtfully such that, you know, probably not that different from how it's evolved in computer architecture, where the CPU is sort of an orchestrator of these really like heavy lifting GPU tasks. That dividing line does shift a little bit, you know, with every generation. And so I think we need to think about knowledge work in that context, like what are human brains good at? What's our silicon brain good at? Let's resegment the work. Let's offload all the stuff that can be automated. Let's go on a hunt for like anything that could save a human CPU cycle. Let's give it to the silicon one. And so I think we're at the early earnings of actually being able to do something about it. Alessio [00:48:00]: It's funny, I gave a talk to a few government people earlier this year with a similar point where we used to make machines to release human labor. And then the kilowatt hour was kind of like the unit for a lot of countries. And now you're doing the same thing with the brain and the data centers are kind of computational power plants, you know, they're kind of on demand tokens. You're on the board of Meta, which is the number one donor of Flops for the open source world. The thing about open source AI is like the model can be open source, but you need to carry a briefcase to actually maybe run a model that is not even that good compared to some of the big ones. How do you think about some of the differences in the open source ethos with like traditional software where it's like really easy to run and act on it versus like models where it's like it might be open source, but like I'm kind of limited, sort of can do with it? Drew [00:48:45]: Yeah, well, I think with every new era of computing, there's sort of a tug of war between is this going to be like an open one or a closed one? And, you know, there's pros and cons to both. It's not like open is always better or open always wins. But, you know, I think you look at how the mobile, like the PC era and the Internet era started out being more on the open side, like it's very modular. Everybody sort of party that everybody could, you know, come to some downsides of that security. But I think, you know, the advent of AI, I think there's a real question, like given the capital intensity of what it takes to train these foundation models, like are we going to live in a world where oligopoly or cartel or all, you know, there's a few companies that have the keys and we're all just like paying them rent. You know, that's one future. Or is it going to be more open and accessible? And I'm like super happy with how that's just I find it exciting on many levels with all the different hats I wear about it. You know, fortunately, you've seen in real life, yeah, even if people aren't bringing GPUs on a plane or something, you've seen like the price performance of these models improve 10 or 100x year over year, which is sort of like many Moore's laws compounded together for a bunch of reasons like that wouldn't have happened without open source. Right. You know, for a lot of same reasons, it's probably better that we can anyone can sort of spin up a website without having to buy an internet information server license like there was some alternative future. So like things are Linux and really good. And there was a good balance of trade to where like people contribute their code and then also benefit from the community returning the favor. I mean, you're seeing that with open source. So you wouldn't see all this like, you know, this flourishing of research and of just sort of the democratization of access to compute without open source. And so I think it's been like phenomenally successful in terms of just moving the ball forward and pretty much anything you care about, I believe, even like safety. You can have a lot more eyes on it and transparency instead of just something is happening. And there was three places with nuclear power plants attached to them. Right. So I think it's it's been awesome to see. And then and again, for like wearing my Dropbox hat, like anybody who's like scaling a service to millions of people, again, I'm probably not using like frontier models for every request. It's, you know, there are a lot of different configurations, mostly with smaller models. And even before you even talk about getting on the device, like, you know, you need this whole kind of constellation of different options. So open source has been great for that. Alessio [00:51:06]: And you were one of the first companies in the cloud repatriation. You kind of brought back all the storage into your own data centers. Where are we in the AI wave for that? I don't think people really care today to bring the models in-house. Like, do you think people will care in the future? Like, especially as you have more small models that you want to control more of the economics? Or are the tokens so subsidized that like it just doesn't matter? It's more like a principle. Yeah. Yeah. Drew [00:51:30]: I mean, I think there's another one where like thinking about the future is a lot easier if you start with the past. So, I mean, there's definitely this like big surge in demand as like there's sort of this FOMO driven bubble of like all of big tech taking their headings and shipping them to Jensen for a couple of years. And then you're like, all right, well, first of all, we've seen this kind of thing before. And in the late 90s with like Fiber, you know, this huge race to like own the internet, own the information superhighway, literally, and then way overbuilt. And then there was this like crash. I don't know to what extent, like maybe it is really different this time. Or, you know, maybe if we create AGI that will sort of solve the rest of the, or we'll just have a different set of things to worry about. But, you know, the simplest way I think about it is like this is sort of a rent not buy phase because, you know, I wouldn't want to be, we're still so early in the maturity, you know, I wouldn't want to be buying like pallets of over like of 286s at a 5x markup when like the 386 and 486 and Pentium and everything are like clearly coming there around the corner. And again, because of open source, there's just been a lot more competition at every layer in the stack. And so product developers are basically beneficiaries of that. You know, the things we can do with the sort of cost estimates I was looking at a year or two ago to like provide different capabilities in the product, you know, cut, right, you know, slashing by 10, 100, 1000x. I think about coming back around. I mean, I think, you know, at some point you have to believe that the sort of supply and demand will even out as it always does. And then there's also like non-NVIDIA stacks like the Grok or Cerebris or some of these custom silicon companies that are super interesting and outperformed NVIDIA stack in terms of latency and things like that. So I guess it'd be a pretty exciting change. I think we're not close to the point where we were with like hard drives or storage when we sort of went back from the public cloud because like there it was like, yeah, the cost curves are super predictable. We know what the cost of a hard drive and a server and, you know, terabyte of bandwidth and all the inputs are going to just keep going down, riding down this cost curve. But to like rely on the public cloud to pass that along is sort of, we need a better strategy than like relying on the kindness of strangers. So we decided to bring that in house and still do, and we still get a lot of advantages. That said, like the public cloud is like scaled and been like a lot more reliable and just good all around than we would have predicted because actually back then we were worried like, is the public cloud going to even scale fast enough to where to keep up with us? But yeah, I think we're in the early innings. It's a little too chaotic right now. So I think renting and not sort of preserving agility is pretty important in times like these. Yeah. Alessio [00:54:01]: We just went to the Cerebrus factory to do an episode there. We saw one of their data centers inside. Yeah. It's kind of like, okay, if this really works, you know, it kind of changes everything. Drew [00:54:13]: And that is one of the things there, like this is one where you could just have these things that just like, okay, there's just like a new kind of piece on the chessboard, like recalc everything. So I think there's still, I mean, this is like not that likely, but I think this is an area where it actually could, you could have these sort of like, you know, and out of nowhere, all of a sudden, you know, everything's different. Yeah. Alessio [00:54:33]: I know one of the management books he references, Ending Growth's, I'm only the paranoid survive. Drew [00:54:37]: Yeah. Alessio [00:54:37]: Maybe if you look at Intel, they did a great job memory to chip, but then it's like maybe CPU to GPU, they kind of missed that thing. Yeah. How do you think about staying relevant for so long now? It's been 17 years you've been doing Dropbox. Drew [00:54:50]: What's the secret? Alessio [00:54:50]: And maybe we can touch on founder mode and all of that. Yeah. Drew [00:54:55]: Well, first, what makes tech exciting and also makes it hard is like, there's no standing still, right? And your customers never are like, oh no, we're good now. They always want more just, and then the ground is shifting under you or it's like, oh yeah, well, files are not even that relevant to the modern. I mean, it's still important, but like, you know, so much is tilted elsewhere. So I think you have to like always be moving and think about on the one level, like what is, and thinking of these different layers of abstraction, like, well, yeah, the technical service we provide is file syncing and storage in the past, but in the future it's going to be different. The way Netflix had to look at, well, technically we mail people physical DVDs and fulfillment centers, and then we have to switch like streaming and codex and bandwidth and data centers. So you, you, you do have to think about that level, but then it's like our, what's the evergreen problem we're solving is an important problem. Can we build the best product? Can we get distribution? Can we get a business model? Can we defend ourselves when we get copied? And then having like some context of like history has always been like one of the reading about the history, not just in tech, but of business or government or sports or military, these things that seem like totally new, you know, and to me would have been like totally new as a 25 year old, like, oh my God, the world's completely different and everything's going to change. You're like, well, there's not a lot of great things about getting older, but you do see like, well, no, this actually has like a million like precedents and you can actually learn a lot from, you know, about like the future of GPUs from like, I don't know how, you know, how formula one teams work or you can draw all these like weird analogies that are super helpful in guiding you from first principles or through a combination of first principles and like past context. But like, you know, build s**t we're really proud of. Like, that's a pretty important first step and really think about like, you sort of become blind to like how technology works as that's just the way it works. And even something like carrying a thumb drive, you're like, well, I'd much rather have a thumb drive than like literally not have my stuff or like have to carry a big external hard drive around. So you're always thinking like, oh, this is awesome. Like I ripped CDs and these like MP3s and these files and folders. This is the best. But then you miss on the other side. You're like, this isn't the end, right? MP3s and folders. It's like an Apple comes along. It's like, this is dumb. You should have like a catalog, artists, playlists, you know, that Spotify is like, Hey, this is dumb. Like you should, why are you buying these things? All the cards, it's the internet. You should have access to everything. And then by the way, why is this like such a single player experience? You should be able to share and they should have, there should be AI curated, et cetera, et cetera. And then a lot of it is also just like drawing, connecting dots between different disciplines, right? So a lot of what we did to make Dropbox successful is like we took a lot of the consumer internet playbook, applied it to business software from a virality and kind of ease of use standpoints. And then, you know, I think there's a lot of, you can draw from the consumer realm and what's worked there and that hasn't been ported over to business, right? So a lot of what we think about is like, yeah, when you sign into Netflix or Spotify or YouTube or any consumer experience, like what do you see? Well, you don't see like a bunch of titles starting with AA, right? You see like this whole, and it went on evolution, right? Like we talked about music and TV went through the same thing, like 10 channels over the air broadcast to 30 channels, a hundred channels, but that's something like a thousand channels. You're like, this has totally lost the plot. So we're sort of in the thousand channels era of productivity tools, which is like, wait, wait, we just need to like rethink the system here and we don't need another thousand channels. We need to redesign the whole experience. And so I think the consumer experiences that are like smart, you know, when you sign into Netflix, it's not like a thousand channels. It's like, here are a bunch of smart defaults. Even if you're a new signup, we don't know anything about you, but because of what the world is watching, here are some, you know, reasonable suggestions. And then it's like, okay, I watched drive to survive. I didn't watch squid game. You know, the next time I sign in, it's like a complete, it's a learning system, right? So a combination of design, machine learning, and just like the courage to like rethink the whole thing. I think that's, that's a pretty reliable recipe. And then you think you're like, all right, there's all that intelligence in the consumer experience. There's no filing things away. Everything's, there's all this sort of auto curated for you and sort of self optimizing. Then you go to work and you're like, there's not even an attempt to incorporate any intelligence or organization anywhere in this experience. And so like, okay, can we do something about that? Alessio [00:58:57]: You know, you're one of the last founder CEOs, like you would talk, then you're like, Toby Lute, some of these folks. Drew [00:59:03]: How, how does that change? I'm like 300 years old and why can't I be a founder CEO? Alessio [00:59:07]: I was saying like when you run, when you run a company, like you've had multiple executives over the years, like how important is that for the founder to be CEO and just say, Hey, look, we're changing the way the company and the strategy works. It's like, we're really taking this seriously versus like you could be a public CEO and be like, Hey, I got my earnings call and like whatever, I just need to focus on getting the right numbers. Like how does that change the culture in the company? Yeah. Drew [00:59:29]: Well, I think it's sort of dovetails with the founder mode whole thing. You know, I think founder mode is kind of this Rorschach test. It's, it's sort of like ill specified. So it's sort of like whatever you, you know, it is whatever you see it. I think it's also like a destination you get to more than like a state of mind. Right. So if you think about, you know, imagine someone, there was something called surgeon mode, you know, given a med student, the scalpel on day one, it's like, okay, hold up. You know, so there's something to be said for like experience and conviction and you know, you're going to do a lot better. A lot of things are a lot easier for me, like 17 years into it than they were one year into it. I think part of why founder mode is so resonant is, or it's like striking such a chord with so many people is, yeah, there's, there's a real power when you have like a directive, intuitive leader who can like decisively take the company like into the future. It's like, how the hell do you get that? Um, and I think every founder who makes it this long, like kind of can't help it, but to learn a lot during that period. And you talk about the, you know, Steve jobs or Elan's of the world, they, they did go through like wandering a period of like wandering in the desert or like nothing was working and they weren't the cool kids. I think you either sort of like unsubscribe or kind of get off the train during that. And I don't blame anyone for doing that. There are many times where I thought about that, but I think at some point you sort of, it all comes together and you sort of start being able to see the matrix. So you've sort of seen enough and learned enough. And as long as you keep your learning rate up, you can kind of surprise yourself in terms of like how capable you can become over a long period. And so I think there's a lot of like founder CEO journey, especially as an engineer. Like, you know, I never like set out to be a CEO. In fact, like the more I like understood in the early days, what CEOs did, the more convinced I was that I was like not the right person actually. And it was only after some like shoving by a previous mentor, like, Hey, don't just, just go try it. And if you don't like it, then you don't have to do it forever. So I think you start founder mode, you're, you're sort of default that because there's like, you realize pretty quickly, like nothing gets done in this company unless the founders are literally doing it by hand, then you scale. And then you're like, you get, you know, a lot of actually pretty good advice that like, you can't do everything yourself. Like you actually do need to hire people and like give them real responsibilities and empower people. And that's like a whole discipline called like management that, you know, we're not figuring out for the first time here, but then you, then there's a tendency to like lean too far back, you know, it's tough. And if you're like a 30 year old and you hire a 45 year old exec from, you know, high-flying company and a guy who was running like a $10 billion P&L and came to work for Dropbox where we were like a fraction of a billion dollar P&L and, you know, what am I going to tell him about sales? Right. And so you sort of recognize pretty quickly, like, I actually don't know a lot about all these different disciplines and like, maybe I should lean back and like let people do their thing. But then you can create this, like, if you lean too far back out, you create this sort of like vacuum, leadership vacuum where people are like, what are we doing? And then, you know, the system kind of like nature reports a vacuum, it builds all these like kind of weird structures just to keep the thing like standing up. And then at some point you learn enough of this that you're like, wait, this is not how this should be designed. And you actually get like the conviction and you learn enough to like know what to do and things like that. And then on the other side, you lean way back in. I think it's more of like a table flipping where you're like, hey, this company is like not running the way I want it. Like something, I don't know what happened, but it's going to be like this now. And I think that that's like an important developmental stage for a founder CEO. And if you can do it right and like make it to that point, like then the job becomes like a lot of fun and exciting and good things happen for the company, good things for happening for your customers. But it's not, it's like a really rough, you know, learning journey. It is. It is. Alessio [01:03:10]: I've had many therapy sessions with founder CEOs. Let's go back to the beginning. Like today, the AI wave is like so big that like a lot of people are kind of scared to jump in the water. And when you started Dropbox, one article said, fortunately, the Dropbox founders are too stupid to know everyone's already tried this. In AI now, it kind of feels the same. You have a lot of companies that sound the same, but like none of them are really working. So obviously the problem is not solved. Do you have any advice for founders trying to navigate like the idea maze today on like what they should do? What are like counterintuitive things maybe to try? Drew [01:03:45]: Well, I think like, you know, bringing together some of what we've covered, I think there's a lot of very common kind of category errors that founders make. One is, you know, I think he's starting from the technology versus starting from like a customer or starting from a use case. And I think every founder has to start with what you know. Like you're, yeah, you know, maybe if you're an engineer, you know how to build a product, but don't know any of the other next, you know, hurdle. You don't know much about the next hurdles you have to go through. So I think, I think the biggest lesson would be you have to keep your personal growth curve out of the company's growth curve. And for me, that meant you have to be like super systematic about training up what you don't know, because no one's going to do that for you. Your investors aren't going to do that. Like literally no one else will do that for you. And so then, then you have to have like, all right, well, and I think the most important, one of the most helpful questions to ask there is like, in five years from now, what do I wish I had been learning today? In three years from now, what do I wish in one year? You know, how will my job be different? How do I work back from that? And so, for example, you know, when I was just starting in 2007, it really was just like coding and talking to customers. And it's sort of like the YC ethos, you know, make something people want and coding and talking to customers are really all you should be doing in that early phase. But then if I were like, all right, well, that's sort of YC phase, what's, what are the next hurdles? Well, a year from now, then I'm going to need, but to get people, we're going to need fundraise, like raise money. Okay. To raise money, we're going to have to like, have to answer all these questions. We have to see like work back from that. And you're like, all right, we need to become like an expert in like venture capital financing. And then, you know, the circle keeps expanding. Then if we have a bunch of money, we're going to need like accountants and lawyers and employees. And I'm not to start managing people. Then two years would be like, well, we're gonna have this like products, but then we're gonna need users. We need money revenue. And then in five years, it'd be like, yeah, we're going to be like tangling with like Microsoft, Google, Apple, Facebook, everybody. And like, somehow we're going to feel like deal with that. And then that's like what the company's got to deal with. And as CEO, I'm going to be responsible for all that. But then like my personal growth, there's all these skills I'm going to need. I'm going to need to know like what marketing is and like what finance is and how to manage people, how to be a leader, whatever that is. And so, and then I think one thing people often do is like, oof, like that it's like imposter syndrome kind of stuff. You're like, oh, it seems so remote or far away that, or I'm not comfortable speaking publicly or I've never managed people before. I haven't this. I haven't been like, and maybe even learning a little bit about it makes it feel even worse. He's like, now I, I thought I didn't know a lot. Now I know I don't know a lot, right. Part of it is more technical. Like how do I learn all these different disciplines and sort of train myself and a lot of that's like reading, you know, having founders or community that are sort of going through the same thing. So that's, that was how I learned. Maybe reading was the single most helpful thing more than any one person or, or talking to people like reading books. But then there's a whole mindset piece of it, which is sort of like, you have to cut yourself a little bit of slack. Like, you know, I wish someone had sort of sat me down and told me like, dude, you may be an engineer, but like, look, all the tech founders that, you know, tech CEOs that you admire, like they actually all, you know, almost all of them started out as engineers, they learned the business stuff on the job. So like, this is actually something that's normal and achievable. You're not like broken for not knowing, you know, no, those people didn't, weren't like, didn't come out of the womb with like shiny hair and Armani suit. You know, you can learn this stuff. So even just like knowing it's learnable and then second, like, but I think there's a big piece of it around like discomfort where it's like, I mean, we're like kind of pushing the edges. I don't know if I want to be CEO or I don't know if I'm ready for this, this, this, like learning to like walk towards that when you want to run away from it. And then lastly, I think, you know, just recognizing the time constant. So five weeks, you're not going to be a great leader or manager or a great public speaker or whatever, you know, think any more than you'll be a great guitar players, you know, play sport that well, or be a surgeon. But in like five years, like actually you can be pretty good at any of those things. Maybe you won't be like fully expert, but you like a lot more latent potential. You know, people have a lot more latent potential than they fully appreciate, but it doesn't happen by itself. You have to like carve out time and really be systematic about unlocking it. Alessio [01:07:36]: How do you think about that for building your team? I know you're a big Pat's fan. Obviously the, that's a great example of building a dynasty on like some building blocks and bringing people into the system. When you're building a company, like how much slack do you have people on, Hey, you're going to learn this versus like, how do you measure like the learning grade of the people you hire? And like, how do you think about picking and choosing? Great question. Drew [01:07:56]: It's hard. Um, what you want is a balance, right? And we've had a lot of success with great leaders who actually grew up with a company, started as an IC engineer or something, then made their way to whatever level our exec team is populated with a lot of those folks. But, but yeah, but there's also a lot of benefit to experience and having seen different environments and kind of been there, done that. And there's a lot of drawbacks to kind of learning by trial and error only. Um, and then even your high potential people like can go up the learning curve faster if they have like someone experienced to learn from now, like experiences in a panacea, either you can, you know, have various organ rejection or misfit or like overfitting from their past experience or cultural mismatches or, you know, you name it, I've seen it all. I've done, I've kind of gotten all the mistake merit badges on that. But I think it's like constructing a team where there's a good balance, like, okay, for the high potential folks who are sort of in the biggest jobs, their lives can, do they either have someone that they're managing them that they can learn from, you know, as a CEO, part of your job or as a manager, like you have to like surround or they help support them. So getting the mentors are getting first time execs like mentors who have been there, done that, or, um, getting them in like, you know, there's usually for any function, there's usually like a social group, like, Oh, chiefs of staff of Silicon Valley. Okay. Like, you know, there's usually these informal kind of communities you can join. And then, um, yeah, you just don't want to be too rotated in one direction or the other, because we've, we've done it. We've like overdone it on the high potential piece, but then like everybody's kind of making dumb mistakes, the bad mistakes are the ones where you're like, either you're making it multiple times or like these are known knowns to the industry, but if they're not known, known, if they're like unknown unknowns to your team, then you're doing, you have a problem. And then again, if you have too much, if you've just only hire external people, like then you're sort of at the mercy, you'll be like whatever random average of whatever culture or practices they bring in can create resentment or like lack of career opportunities. Um, so it's really about how do you get, you know, it doesn't really matter if it's like exactly 50 50, I don't think about a sort of perfect balance, but you just need to be sort of tending that garden continuously. Awesome. Alessio [01:09:57]: Drew, just to wrap, do you have any call to actions? Like who should come work at Dropbox? Like who should use Dropbox? Anything you want, uh, you want to tell people? Drew [01:10:06]: Well, I'm super, I mean, today's a super exciting day for, cause we just launched dash for business and, you know, we've talked a little bit about the product. It's like universal search, universal access control, a lot of rethinking, sharing for the modern environment. But you know, what's personally exciting, you could talk about the product, but like the, it's just really exciting for me to like, yeah, this is like the first, like most major and most public step we've taken from our kind of Dropbox 1.0 roots. And there's probably a lot of people out there who either like grew up not using Dropbox or like, yeah, I used Dropbox like 10 years ago and it was cool, but I don't do that much of fun. So I think there's a lot of new reasons to kind of tune into what we're doing. And, and it's a lot of, it's been a lot of fun to, I think like the sort of the AI era has created all these new like paths forward for Dropbox that wouldn't have been here five years ago. And then, yeah, to the founders, like, you know, hang in there, do some reading and don't be too stressed about it. So we're pretty lucky to get to do what we do. Yeah. Alessio [01:11:05]: Watch the Pats documentary on Apple TV. Drew [01:11:08]: Yeah, Bill Belichick. I'm still Pats fan. Really got an F1. So we're technology partners with McLaren. They're doing super well. Alessio [01:11:15]: So were you a McLaren fan before you were technology partner? So did you become partners? Drew [01:11:19]: It's sort of like co-evolved. Yeah. I mean, I was a fan beforehand, but I'm like a lot more of a fan now, as you'd imagine. Alessio [01:11:24]: Awesome. Well, thank you so much for the time, Drew. This was great. It was a lot of fun. Drew [01:11:28]: Thanks for having me. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.latent.space/subscribe https://www.latent.space/subscribe?utm_medium=podcast&utm_campaign=CTA_2

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:04Alessio Fanelli?

  Hey everyone, welcome to the Late in Space Podcast. This is Salasio, Partner and CTO at Decibel Partners. And there's no switch today, but I'm joined by Drew Houston of Dropbox. Welcome Drew.

• 0:13Speaker 01

  Thanks for having me.

• 0:14Alessio Fanelli?

  So we're not going to talk about the Dropbox story. We're not going to talk about the Chinatown bus and the flash drive and all of that. I think you talked enough about it.

• 0:22Speaker 01

  Yep.

• 0:22Alessio Fanelli?

  Where I want to start is you as an AI engineer. So as you know, most of our audience is engineering folks, kind of like technology leaders. You obviously run Dropbox, which is a huge company, but you also do a lot of coding. I think I saw you spend almost 400 hours just like coding. So let's start there.

• 0:37Speaker 01

  Sure.

• 0:37Alessio Fanelli?

  What was like the first interaction you had with like an LLM API? And when did the journey start for you?

• 0:43Speaker 01

  Yeah, well, I think probably like all AI engineers or whatever you call an AI engineer, those people started out as engineers before that. So like engineering is my first love. I mean, I grew up as a little kid. I was that kid. My first line of code was at five years old. Just really loved, I want to make computer games, like all this whole path. That also led me into startups and eventually starting Dropbox. And then with AI specifically, I studied computer science. I got my, I did my undergrad, but I didn't do like grad level computer science. I didn't, I sort of got distracted by all the startup thing. So I didn't do grad level work, but about several years ago, I mean, a couple of things. So one is I sort of, I knew I wanted to go from being an engineer to a founder. And then, but sort of the becoming a CEO part was sort of backed into the job. And so a couple of realizations.

• 1:28Speaker 01

  One is that, I mean, there's a lot of like repetitive and like manual work you have to do as an executive that is actually lends itself pretty well to automation, both for like my own convenience and then out of interest in learning, I guess what we call like classical machine learning these days. I started really trying to wrap my head around understanding machine learning and informational retrieval more, more formally. So I'd say maybe 2016, 2017, started writing these successively, more elaborate scripts to like understand basic like classifiers and regression and again, like basic information retrieval and NLP back in those days. And there are sort of like two things that came out of that.

• 2:09Speaker 01

  One is techniques are super powerful and even just like studying like old school machine learning was a pretty big inversion of the way I had learned engineering, right? You know, I started programming, when everyone starts programming, you're, you're sort of the human, you're giving an algorithm to the, and spelling out to the computer how it should run it. And then machine learning, here's machine learning where it's like actually flip that like give it sort of the answer you want and it'll figure out the algorithm, which was pretty mind bending. And it was both like pretty powerful when I would write tools like figure out like time audits or like where's my time going? Is this meeting a one on one? Or is it a recruiting thing? Or is it a product strategy thing? I started out doing that manually with my assistant, but then found that this was like

• 2:53Speaker 01

  a very like automatable task. And so which also had the side effect of teaching me a lot about machine learning. But then there was this big problem, like anytime you it was very good at like tabular structured data, but like anytime it hit, you know, the usual malformed English that humans speak, it would just like fall over. I had to kind of abandon a lot of the things that I wanted to build because like there's no way to like parse text, like maybe it would sort of identify the part of speech in a sentence or something. But then fast forward to the L L M. I mean, actually, I started trying something like this, what we would call like very small L L Ms before kind of the G P T class models. And it was like super hard to get that those things working. So like these 500 parameter models would just be like hallucinating and repeating and, you

• 3:34Speaker 01

  know. So actually, I'd kind of like written it off a little bit. But then the chat G P T launch and G P T three for sure. And then once people figured out like prompting and instruction tuning, this was sort of like November ish, 2022, like everybody else, sort of had the chat G P T launch being the starting gun for the whole A I era of computing. And then having A P I access to three and then early access G P T four. I was like, oh, man, it's happening. And so I was literally on my honeymoon and we're like on a beach in Thailand and I'm like coding these like A I tools to automate like writing or to assist with writing and all these different. You're like a never go back to work. I'm gonna automate all of it before I get it's fine.

• 4:17Speaker 01

  And I was just, you know, ever since then, I mean, I've always been like coding, like prototypes and just stuff to make my life more convenient. But like escalated a lot after 22. And yeah, I spent, I checked, I think it was probably like over 400 hours this year so far coding because I had my paternity leave where I was able to work on some special projects. But yeah, it's a super important part of like my whole learning journey is like being really hands on with these things. And I mean, it's probably not a typical recipe, but I really love to get down to the metal. As far as how this stuff works.

• 4:47Alessio Fanelli?

  Yeah, Swyx and I were with Sam Altman October 22. We were like at a hack day at OpenAI. And that's why we started this podcast eventually. But you did an interview with Sam like seven years ago. And he asked you, what's the biggest opportunity in startups? And you were like, machine learning and AI. And you were almost like too early, right? It's like maybe seven years ago, the models weren't quite there. How should people think about revalidating like expectations of this technology? You know, I think even today, people will tell you, oh, models are not really good at X because they were not good 12 months ago, but they're good today. What's your process?

• 5:19Speaker 01

  Heuristics for thinking about that or how is? Yeah, I think the way I look at it now is pretty, has evolved a lot since when I started. I mean, I think everybody intuitively starts with like, all right, let's try to predict the future or imagine like, what's this great end state we're going to get to? And the tricky thing is like, often those prognostications are right. But they're right in terms of direction, but not when. For example, you know, even in the early days of the internet, 90s, when things were even like tech space and you know, even before like the browser or things like that. People are like, oh, man, you're going to have, you know, you're going to be able to order food. You get like a Snickers delivered to your house. You're going to be able to watch any movie ever created. And they were right, but they were like, you know, it took, you know, 20 years for that to actually happen.

• 6:01Speaker 01

  And before you got to DoorDash, you had to get, you started with like Webvan and Cosmo. And before you get to Spotify, you had to do like Napster and Kazaa and LimeWire. And like a bunch of like Brooke and Britney Spears MP3s and malware. So I think the big lesson is being early is the same as being wrong. Being late is the same as being wrong. So really, how do you calibrate timing? And then I think with AI, it's the same thing. And people are like, oh, it's going to completely upend society in all these positive and negative ways. I think that's like, most of those things are going to come true. The question is like, when is that going to happen? And then with AI specifically, I think there's also a, in addition, sort of the general tech category error of like jumping too fast to the future. I think that AI is particularly susceptible to that. And you look at self-driving, right?

• 6:46Speaker 01

  This idea of like, oh, my God, you can have a self-driving car. Captured everybody's imaginations 10, 12 years ago. And you know, people are like, oh, man, in two years, there's not going to be another year. There's not going to be a human driver on the road to be seen. It didn't work out that way, right? We're still 10, 12 years later, where we're in a world where you can sort of sometimes get a Waymo in like one city on earth. Exciting, but just took a lot longer than. People think and the reason is there's a lot of like time, there's like a lot of engineering challenges, but then there's a lot of other like societal time constants that are hard to compress. So one thing I think what you can learn from things like self-driving is they have these levels of autonomy. That's a useful kind of framework in driving or these like maturity levels. People sort of skip to like level five, full autonomy,

• 7:28Speaker 01

  or we're going to have like an autonomous knowledge worker that's going to take, that's going to, and then we won't need humans anymore, kind of projection that that's going to take a long time. But then when you think about level one or level two, like these little assistive experiences, you know, we're seeing a lot of traction with those. So what you see really working is the level one autonomy in the AI world would be like the tab autocomplete and copilot, right? Get up copilot and then maybe a little higher is like the chatbot type interface. Obviously, you want to get to the highest level you can to build a good product, but the reliability just isn't and the capability just isn't there in the early innings. And so, and then you think of other level one, level two type things like Google Maps probably did more for self-driving than than literal self-driving.

• 8:13Speaker 01

  Like a billion people have like the ability to have like maps and navigation just like taken care of for you autonomously. So I think the timing and maturity are really important factors to include.

• 8:23Alessio Fanelli?

  The thing with self-driving, maybe one of the big breakthroughs was like simulation. So it's like, okay, instead of driving, we can simulate these environments. It's really hard to do with knowledge work, you know? How do you simulate like a product review? How do you simulate these things? I'm curious if you've done any experiments. I know some companies have started to build kind of like a virtual personas that you can like bounce ideas off of.

• 8:41Speaker 01

  I mean, fortunately, in a company, you generate lots of, you know, actual human training data all the time. And then I also just like start with myself, like, all right, I can tree out, you know, it's pretty tricky even within your company to be like, all right, let's open all this up as quote, training data. But, you know, I can start with my own emails or my own calendar or own stuff without running into the same kind of like privacy or other concerns. So I often like start with my own stuff. And so that is like a one level of bootstrapping. But actually, four or five years ago, during COVID, we decided, you know, a lot of companies were thinking about how do we go back to work? We decided Really lean into remote and distributed work because I thought, you know, this is going to be the biggest change to the way we work in our lifetimes.

• 9:25Speaker 01

  And COVID kind of ripped up a bunch of things, but I think everybody was sort of pleasantly surprised how with a lot of knowledge work, you could just keep going. And actually, you were sort of fine. Work was decoupled from your physical environment, from being in a physical place, which meant that things people had dreamed about since the 50s or 60s, like telework, like you actually could work from anywhere, and that was now possible. So we decided to really lean into that because we debated, should we sort of hit the fast forward button or should we hit the rewind button, go back to 2019? You know, obviously, that's been playing out over the last few years. And we decided to basically turn, we went like 90% remote. We still, the in-person part's really important. We can kind of come back to our working model, but we're like, yeah, this is, everybody is going to be in some kind of like distributed or hybrid state. So like,

• 10:08Speaker 01

  instead of like running away from this, like, let's do a full send. Let's really go into it. Let's live in the future a few years before our customers. Let's like turn Dropbox into a lab for distributed work. And we do that like quite literally, both of the working model and then increasingly with our products. And then, absolutely, like we have products like Dropbox Dash, sorry, Universal Search products. That was like very elevated in priority for me after COVID because like now you have, we're putting a lot more stress on the system and on our screens. It's a lot more chaotic and overwhelming. And so even just like getting the right information, the right person at the right time is a big fundamental challenge in knowledge work in these, in the distributed world, like big problem today is still getting, you know, has been getting bigger.

• 10:51Speaker 01

  And then for a lot of these other workflows, yeah, there's, we can both get a lot of natural like training data from just our own like strategy docs and processes. There's obviously a lot you can do with synthetic data and, you know, actually like LMS are pretty good at being like imitating generic knowledge workers. So it's, it's kind of funny that way. But yeah, the way I look at it is like really turn Dropbox into a lab for distributed work. You think about things, like what are the big problems we're going to have? It's just the complexity on our screens just keeps growing and and the whole environment gets kind of more out of sync with what makes us like cognitively productive and engaged. And then even something like Dash was initially seated. I made a little personal search engine because I was just like personally frustrated with

• 11:33Speaker 01

  not being able to find my stuff. And along that whole learning journey with AI, like the vector search or semantic search, like that had just been the tooling for that. The open source stuff had finally gotten to a place where it was a pretty good developer experience. And so You know, in a few days, I had sort of a hello world type search engine. And I'm like, oh, my God, like this completely works. You don't even have to get the keywords right. The relevance and ranking is super good. We even like untuned. So I guess that's to say like, I've been surprised by if you choose like the right algorithm and the right approach, you can actually get like super good results without having like a ton of data. And even with LLMs, you can apply all these other techniques to give them to kind of bootstrap, kind of like task maturity pretty quickly.

• 12:14Alessio Fanelli?

  Before we jump into the dash, let's talk about the Drew Houston AI engineering. So IDE, let's write that down. What IDE do you use? Do you use Cursor, VS Code? Do you use any coding assistant like with chat? Is it just autocomplete?

• 12:27Speaker 01

  Yeah, yeah, both. So I use VS Code as like my daily driver, although I'm like super excited about things like Cursor or the AI agents. I have my own like stack underneath that. I mean, some off the shelf parts, some pretty custom. So I use the continue dev, just like AI chat UI basically as just the UI layer. But I also proxy the request. I don't, I proxy the request to my own backend, which is sort of like a router. You can use any backend. I mean, Sonar 3.5 is probably the best all around. But then these things are like pretty limited if you don't give them the right context. And so part of what the proxy does is like, there's a separate thing where I can say like include all these files by default with the requests. Then it becomes a lot easier and like without like cutting and pasting. And I'm building mostly like prototype toy apps.

• 13:11Speaker 01

  So you like a front end react thing and a Python backend thing. And so it can do these like end to end diffs basically. And then I also like love being able to host everything locally or do it offline. So I have my own, when I'm on a plane or something or where like you don't have access or the internet's not reliable, I actually bring a gaming laptop on the plane with me. It's like little like blue briefcase looking thing. And then I literally hook up a GPU like into one of the outlets. And then I have, I can do like transcription, I can do like autocomplete. Like I have an eight billion Llama will run fine.

• 13:44Alessio Fanelli?

  And you're using like a Ollama to run the model?

• 13:47Speaker 01

  No, I use, I have my own like LLM inference stack. I mean, I uses the backend somewhat interchangeable. So everything from like X Llama to VLM or SG Lang. There's a bunch of these different backends you can use. And then I started like working on this stuff before all this tooling was like really available. So, you know, over the last several years, I've built like my own like whole crazy environment and like in stack here. So I'm a little nuts about it.

• 14:11Alessio Fanelli?

  Yeah. What's the state of the art for, I guess not state of the art, but like when it comes to like frameworks and things like that? Do you like using them? I think maybe a lot of people say, hey, things change so quickly. They're like trying to abstract things. It's maybe too early today.

• 14:24Speaker 01

  As much as I do a lot of coding, I have to be pretty surgical with my time. I don't have that much time, which means I have to sort of like scope my innovation to like very specific places or like my time. So for the front end, it'll be like a pretty vanilla stack, like a, you know, next JS, React based thing. And then these are toy apps. So it's like Python, Flask, SQLite, and then all the different, then there's a whole other thing on like the backend. Like, how do you get sort of run all these models locally or with a local GPU? The scaffolding on the front end is pretty straightforward. The scaffolding on the backend is pretty straightforward. But then a lot of it is just like the LLM inference and control over like fine grain aspects of how you do generation, caching, things like that.

• 15:05Speaker 01

  And then there's a lot, like a lot of the work is how do you take sort of go to an IMAP, like take an email, get a new or a document or a spreadsheet or, you know, any of these kinds of primitives that you work with, and then translate them, you know, render them in a format that an LLM kind of understands. So there's like a lot of work that goes into that too.

• 15:23Alessio Fanelli?

  Yeah. Yeah, I built a kind of like email three I assistant. And like, I would say 80% of the code is like Google and like pulling the emails. And then the actual AI part is pretty easy.

• 15:33Speaker 01

  Yeah, and even same experience. And then I tried to do all these like NLP things. And then to my dismay, like a bunch of reg Xs were like, got you like 95% of the way there. So I still leave it running. I just haven't really built like the LLM, LLM powered version of it yet.

• 15:50Alessio Fanelli?

  Yeah. Any thoughts on RAG versus long context, especially with Dropbox? Do you just want to shove things in? Have you seen that be a lot better?

• 15:59Speaker 01

  Well, they kind of have different strengths and weaknesses. So you need both for different use cases. I mean, it's been awesome in the last 12 months. Like now you have these long context models that can actually do a lot. You can put a book in Sonnet's context. And then now with the later versions of Llama, you can have 128K context. So that's sort of the new normal, which is awesome. And that wasn't even the case a year ago. That said, models don't always use... And certainly, like, local models don't use the full context well, fully yet. And actually, if you provide too much irrelevant context, the quality degrades a lot. And so I say, in the open source world, we're still just getting to the cusp of like, the full context is usable. And then, of course, when you're something like Dropbox Dash,

• 16:42Speaker 01

  it's basically building this whole brain that's read everything your company's ever written. And so that's not going to fit into your context window. So you need RAG, just as a practical reality. And for a lot of similar reasons, you need RAM and hard disk in conventional computer architecture. And I think these things will keep horse trading. Like, maybe if a million or 10 million is the new tokens, is the new context length, maybe that shifts. Maybe the bigger picture is like, it's super exciting to talk about the LLM and that piece of the puzzle. But there's this whole other scaffolding of more conventional retrieval or conventional machine learning. Especially if you have to scale up products to millions of people, you do, and your toy app is not going to scale to that from a cost or latency or performance standpoint.

• 17:25Speaker 01

  So I think you really need these hybrid architectures where you have very purpose-fit tools. You're probably not using Sonnet 3.5 for all of your normal product use cases. You're going to use a fine-tuned 8 billion model or sort of the minimum model that gets you the right output. And then a smaller model also has a lot more cost and latency. There's much better characteristics on that front.

• 17:48Alessio Fanelli?

  Yeah, let's jump into the Dropbox AI story. So

• 17:50Speaker 01

  Sure.

• 17:51Alessio Fanelli?

  Your initial prototype was Files GPT.

• 17:54Speaker 01

  Yeah.

• 17:55Alessio Fanelli?

  How did it start? And then how did you communicate that internally? You know, I know you have a pretty strong like memo culture.

• 18:02Speaker 01

  Sure.

• 18:02Alessio Fanelli?

  When were you like, okay, hey, we gotta really take this seriously?

• 18:05Speaker 01

  Yeah. Well, on the latter, it was, so how do we, I mean, I'll say like how we took Dropbox, how AI seriously as a company started kind of around that honeymoon time, unfortunately. In January, I wrote this like memo to the company like around basically like how we need to play offense in 23 and that most of the time, the kind of concrete is set and like the winners of the winners and things are kind of frozen. But then with these new eras of computing, like the PC or the internet or the phone or the concrete unfreezes and you can sort of build, do things differently and have a new set of winners. It's sort of like a new season starts. As a result of a lot of that sort of personal hacking and just like thinking about this, I'm like, yeah, this is an inflection point in the industry.

• 18:49Speaker 01

  Like, we really need to change how we think about our strategy. And then becoming an AI-first company was probably the headline thing that we did. And then, and then that got, and then calling on everybody in the company, really think about in your world, how is AI going to reshape your workflows? Or what's sort of the AI native way of thinking about your job? And you know, File GPT, which is sort of this Dropbox AI kind of initial concept, that actually came from our engineering team as you know, as we like called on everybody, like really think about what we should be doing that's new or different. So it's kind of organic and bottoms up, like a bunch of engineers just kind of hacked that together. And then that materializes basically when you preview a file on Dropbox, you can have kind of the most straightforward possible integration

• 19:32Speaker 01

  of AI, which is a good thing. Like basically you have a long PDF, you want to be able to ask questions of it. And so like a pretty basic implementation of RAG and being able to do that when you preview a file on Dropbox. So that was the origin of that. That was like back in 2023, when we really just like the starting engines had just, you know, gone, gone.

• 19:53Alessio Fanelli?

  It's funny where you're busy like these files that people have, they really don't want them in a way. You know, like you're storing all these files and like you actually don't want to interact with them. You want a layer on top of it.

• 20:02Speaker 01

  Yeah.

• 20:02Alessio Fanelli?

  And that's kind of what also takes you to a dash eventually, which is like, hey, you actually don't really care where the file is. You just want to be the place that aggregates it. How do you think about what people will know about files? You know, are files the actual file? Are files like the metadata and they're just kind of like a pointer that goes somewhere? And you don't really care what it is? Yeah. And you talked about-

• 20:22Speaker 01

  Totally. Yeah. I mean, there's a lot of potential complexity in that question, right? Or is it a, you know, what's the difference between a file and a URL? And you could go into the technicals, which is always it's like pass by value, pass by reference. Okay. What's the format like? All right. Well, now if it's real time collaborative, it's not really a flat file. It's like a structured data, you're sort of collaborative, that's keeping in sync, blah, blah, blah. I actually don't start there at all. I just start with like, what do people, like what do humans? Let's work back from like how humans think about this stuff or how they should think about this stuff. Meaning like, I don't think about, oh, here are my files and here are my links or a cloud docs. I'm just sort of like, oh, here's my stuff. This, this, here's sort of my documents. Here's my media. Here's my projects. Here are the people I'm working with. So it starts from primitives more like those. Like how do people, how do humans think about these things?

• 21:07Speaker 01

  And then, then start from like a more ideal experience. Cause if you think about it, we kind of have this situation that will look like particularly medieval in hindsight where, all right, how do you manage your work stuff? Well, on a lot, you know, on one side of your screen, you have this file browser that literally hasn't changed since the early eighties. Right. You could take someone from the original Mac and sit them in front of like a computer and they'd be like, this is it. And that's, it's been 40 years.

• 21:30Alessio Fanelli?

  Right.

• 21:31Speaker 01

  Right. Then on the other side of your screen, you have like Chrome or a browser that has so many tabs open. You can no longer see text or titles. This is the state of the art for how we manage stuff at work. Interestingly, neither of those experiences was purpose built to be like the home for your work stuff or even anything related to it like, and so- It's important to remember we get stuck in these local maxima pretty often in tech. We're obviously aware that files are not going away, especially in certain domains. Like the format really matters and where files are still going to be the tool you use for like, if there's something big, right? If you have a big video file, that kind of format in a file makes sense.

• 22:14Speaker 01

  There's a bunch of industries where it's like construction or architecture or sort of these domain-specific areas. You know, media generally, if you're making music or photos or video, that all kind of fits in the big file zone where Dropbox is really strong and that's like what customers love us for. But you know, it's also pretty obvious that a lot of stuff that used to be in Word docs or Excel files, like all that has tilted towards the browser and that tilt is going to continue. So with Dash, we wanted to make something that was really like cloud-native, AI-native, and deliberately like not be tied down to the abstractions of the file system. Now, on the other hand, it would be like ironic and bad if we then like fractured the experience such like, well, if it touches a file, it's a syncing metaphor and this app and if it's

• 22:59Speaker 01

  a URL, it's like this completely different interface. So there's a convergence that I think makes sense over time. But you know, but I think you have to start from like not so much the technology, start from like what do the humans want? And then like, what's the idealized product experience? And then like, what are the technical underpinnings of that that can make that good experience?

• 23:20Alessio Fanelli?

  I think it's counterintuitive that in Dash, you can connect Google Drive, right? Because you think about Dropbox, it's like, well, it's file storage. You really don't want people to store files somewhere, but the reality is that they do. How do you think about the importance of storage? And like, do you kind of feel storage is like almost solved? Where it's like, hey, you can kind of store these files anywhere. What matters is like access.

• 23:38Speaker 01

  It's a little bit nuanced in that if you're dealing with like large quantities of data, it actually does matter. The implementation matters a lot or like if you're dealing with like, you know, 10 gig video files like that, then you sort of inherit all the problems of sync and have to go into a lot of the challenges that we've solved. Touching on a pretty important question, like what is the value we provide? What does Dropbox do? And probably like most people, I would have said like, well, Dropbox syncs your files.

• 24:03Speaker 01

  And we didn't even really have a mission of the company in the beginning. I'm just like, yeah, I just don't want to carry a thumb drive around and be... Life would be a lot better if our stuff just like lived in the cloud. And I just didn't have to think about like, what device is the thing on? Or what operating, why are these operating systems fighting with each other and incompatible? You know, I just want to abstract all of that away. But then, so we thought, even we were like, all right, Dropbox provides storage. But when we talked to our customers, they're like, that's not how we see this at all. Like, actually, Dropbox is not just like a hard drive in the cloud. It's like the place where I go to work. Or it's a place like I started a small business is a place where my dreams come true. Or it's like, yeah, it's not keeping files in sync. It's keeping people in sync. It's keeping my team in sync. And so they're using this kind of language where we're like, wait, okay.

• 24:47Speaker 01

  Yeah, because I, you know, I don't know, storage probably is a commodity or what we do is a commodity. Like, but then we talked to our customers like, no, like, we're not buying the storage. We're buying like the ability to access all of our stuff in one place. We're buying the ability to share everything and sort of, in a lot of ways, people are buying the ability to work from anywhere. And Dropbox was kind of, the fact that it was like file syncing was an implementation detail of this higher order need that they had. So, you know, I think that's where we start too. It's just like, what is the sort of higher order thing, the job the customer's hiring Dropbox to do? Storage in the new world is kind of incidental to that. I mean, it still matters for things like video or those kinds of workflows. The value of Dropbox had never been, we provide you like the cheapest bits in the cloud.

• 25:28Speaker 01

  But it is a big pivot from Dropbox is the company that syncs your files to now where we're going is Dropbox is the company that kind of helps you organize all your cloud content. I started the company because I kept forgetting my thumb drive. But with the question I was really asking was like, why is it so hard to like find my stuff, organize my stuff, share my stuff, keep my stuff safe? You know, I'm always like one washing machine. I would leave like my little thumb drive with all my prior company stuff on, in the pocket of my shorts and then almost wash it and destroy it. And so I was like, why do we have to, this is like medieval that we have to think about this. So that's, that same mindset is how I approach where we're going. But I think, and unfortunately, the, we're sort of back to the same problems. Like, it's really hard to find my stuff.

• 26:13Speaker 01

  It's really hard to organize myself. It's hard to share my stuff. It's hard to secure my content at work. Now, the problem is the same, the shape of the problem and the shape of the solution is pretty different. You know, instead of 100 files on your desktop, it's now 100 tabs in your browser, et cetera. But I think that's the starting point.

• 26:30Alessio Fanelli?

  How has the idea of a product evolved for you? So, you know, famously Steve Jobs started by Dropbox and he's like, you know, this is just a feature, it's not a product. And then you build like a $10 billion feature.

• 26:39Speaker 01

  Yeah.

• 26:40Alessio Fanelli?

  How, in the age of AI, how do you think about, you know, maybe things that used to be a product are now features because the AI on top of it? It's like the product, like what's your mental model to think about it?

• 26:50Speaker 01

  Yeah. So I don't think there's really like a bright line. I don't know if like I use the word features and products in my mental model that much of how I break it down. Cause this is kind of a, it's a good question. I mean, I don't not think about features or not think about products, but it does start from that place of like, all right, we have all these new colors we can paint with. And all right, what are these higher order needs that are sort of evergreen, right? So people will always have stuff at work. They're always need to be able to find it or all the verbs I just mentioned. It's like, okay, how can we make like a better painting and how can we, and then how can we use some of these new colors? And then, yeah, it's like pretty clear that after the large models, the way you find stuff, organize stuff, share stuff, it's going to be completely different.

• 27:34Speaker 01

  After COVID, it's going to be completely different. So that's the starting point, but I think it is also important to, you know, you have to do more than just work back from the customer and like what they're trying to do. Like you have to think about, and we've learned a lot of this the hard way sometimes. You might start with a customer, you might start with a job to be on, then you're like, all right, what's the solution to their problem? Or like, can we build the best product that solves that problem, right? So can we build the best way to find your stuff in the modern world? Like, well, yeah, right now the status quo for the vast majority of the billion knowledge workers is they have like 10 search boxes at work that each search 10% of your stuff. Like that's clearly broken. Obviously you should just have like one search box, all right, so we can do that. And that by also has to be like, I'll come back to defensibility in a second.

• 28:19Speaker 01

  But like, can we build the right solution that is like meaningfully better from the status quo? Like, yes, clearly. Okay. Then can we like get distribution and growth? Like that's sort of the next thing you learned is, as a founder, you start with like, what's the product? What's the product? Then you're like, wait, wait, we need distribution and we need a business model. So those are the next kind of two dominos you have to knock down or sort of needles you have to thread at the same time. So, all right, how do we grow? I mean, Dropbox 1.0 is really this like self-serve viral model that is a lot of, we sort of took, borrowed from a lot of the. Consumer internet playbook and like what Facebook and social media were doing and then translated that to sort of the business world. But how do you get distribution, especially as a startup? And then a business model like, all right, storage happened to be some, in the beginning, happened to be something people were willing to pay for.

• 29:02Speaker 01

  They recognized that, you know, okay, if I don't buy something like Dropbox, I'm gonna have to buy an external hard drive, I'm gonna have to buy a thumb drive, I'm gonna have to pay for something one way or another. People were already paying for things like backup, so we felt good about that. But then the last domino is like defensibility. Okay, so you build this product, you get this business model, but then, you know, what do you do in the incumbents? The next chess move for them is I just like copy bundle kill. So they're gonna copy your product, they'll bundle it with their platforms and they'll like give it away for free or no added cost. And, you know, we had a lot of, you know, scar tissue from being on the wrong side of that. Now, you don't need to solve all four, for all four or five variables or whatever at once, or you can sort of have, you know, some flexibility, but the more of those gates

• 29:45Speaker 01

  that you get through, you sort of add a 10x to your valuation. And so with AI, I think, you know, there's been a lot of focus on the large language model, but it's like, large language models are a pretty bad business from a, you know, if you sort of take off your tech lens and your sort of business lens, like, they're sort of this weirdly self-commoditizing thing where, you know, models only have value if they're kind of on this like Pareto frontier of size and quality and cost. Being number two, you know, if you're not on that frontier, the second, the second the frontier moves out, which it moves out every week, like, your model literally has zero economic value because it's dominated by the new thing. LLMs generate output that can be used to train or improve, so there's this weird, peculiar things that are specific to the large language model.

• 30:30Speaker 01

  And then you have to like be like, all right, where's the value gonna accrue in the stack or the value chain? And, you know, certainly at the bottom with Nvidia and the semiconductor companies. And then, then it's gonna be at the top, like the people who have the customer relationship, who have the application layer. Those are a few of the lenses that I look at a question like that through.

• 30:48Alessio Fanelli?

  Do you think AI is making people more careful about sharing their data at all? People are like, oh, data is important, but it's like, whatever, I'm just throwing it out there. Now, everybody's like, but are you gonna train on my data? And like, your data is actually not that good to train on anyway. But like, how have you seen, especially customers, like, think about what to put in, what to not?

• 31:06Speaker 01

  I mean, everybody should be, well, if everybody is concerned about this, and everybody should be concerned about this, right? Because nobody wants their personal, or company's information, to be kind of ground up into little pellets to like, sell you ads, or train the next foundation model. I think it's massively top of mind for every one of our customers, and me personally, and with my Dropbox hat on, it's like, so fundamental. And, you know, we had experience with this too, with Dropbox 1.0, the same kind of resistance. Like, wait, I'm gonna take my stuff on my hard drive and put it on your server somewhere. Are you serious? What could possibly go wrong? And, you know, before that, I was like, wait, you're gonna sell me? I'm gonna put my credit card number into this website? And before, I was like, hey, I'm gonna take all my cash and put it in a bank instead of under my mattress?

• 31:51Speaker 01

  You know, so there's a long history of like, tech and comfort. So, in some sense, AI is kind of another round of the same thing, but the issues are real. And then when I think about like, defensibility for Dropbox, like, that's actually a big advantage that we have. Is one, our incentives are very aligned with our customers, right? We only get, we only make money if you pay us, and you only pay us if we do a good job. So, we don't have any like, side hustle, you know, but we're not training the next foundation model. You know, we're not trying to sell you ads. Actually, we're not even trying to lock you into an ecosystem. Like, the whole point of Dropbox is it works everywhere. Because I think one of the big questions, and we've circling around this, we're like, in the world of AI, where should our lane be? Like, every startup has to ask, or in every big company has to ask, like, where can we really win?

• 32:33Speaker 01

  But, to me, it was like, a lot of the like, trust advantages, about being platform agnostic, having like, a very clean business model, not having these other incentives. And then, we also are like, super transparent. We were transparent early on, we're like, alright, we're gonna establish these AI principles. Very table-stake stuff of like, here's transparency, we wanna give people control, we wanna cover privacy, safety, bias, like, fairness, all these things. And we put that out up front, to put some sort of explicit guardrails out, where like, hey, we're, you know, because everybody wants like, a trusted partner, as they sort of go into the wild world of AI. And then, you also see people cutting corners, and, you know, or just there's a lot of uncertainty, or, you know, moving the pieces around after the fact, which no one feels good about.

• 33:14Alessio Fanelli?

  I mean, I would say the last 10, 15 years, the race was kind of being the system of record, being the storage provider. I think today it's almost like, hey, if I can use Dash to like access my Google Drive app, why would I pay Google for like their AI features? So like vice versa, you know, if I can connect my Dropbox storage to this other AI assistant. How do you kind of think about that? About, you know, not being able to capture all the value and how open people will stay? I think today things are still pretty open, but I'm curious if you think things will get more closed or like more open later.

• 33:42Speaker 01

  Yeah, well, I think you have to get the value exchange right. And I think you have to be like a trustworthy partner. Or like no one's going to partner you with you if they think you're going to eat their lunch, right? Or if you're going to disintermediate them. And like all the companies are quite sophisticated with how they think about that. So we try to, like, we know that's going to be the reality. So we're actually not trying to eat anyone's like Google Drive's lunch or anything. Actually, we'll like integrate with Google Drive, we'll integrate with OneDrive, really any of the content platforms, even if they compete with file syncing. So that's actually a big strategic shift. We're not really reliant on being like the store of record. And there are pros and cons to this decision. But if you think about it, we're basically like providing all these apps more engagement. We're like helping users do what they're really trying to do, which is to get, you know, that Google Doc or whatever.

• 34:25Speaker 01

  And we're not trying to be like, oh, by the way, use this other thing. This is all part of our like brand reputation. It's like, no, we give people freedom to use whatever tools or operating system they want. We're not taking anything away from our partners. We're actually like making it, making their thing more useful or routing people to those things. I mean, on the margin, there was something like, well, okay, to the extent you do rag and summarize things, maybe that doesn't generate a click. Okay. You know, we also know there's like infinity investment going into like the work agents. So we're not really building like a Copilot or Gemini competitor. Not because we don't like those, we don't find that thing like captivating. Yeah, of course. But just like, you know, you learn after some time in this business that like, yeah, there's some places that are just going to be such kind of red oceans or just like super big battlefields.

• 35:10Speaker 01

  Everybody's kind of trying to solve the same problem and they just start duplicating all each other's effort. And then meanwhile, you know, I think the concern would be is like, well, there's all these other problems that aren't being properly addressed by AI. And I was concerned that like, yeah, and everybody's like fixated on the agent. Or the chatbot interface, but forgetting that like, hey, guys, like, we have the opportunity to like, really fix search, or build a self-organizing Dropbox or environments, or there's all these other things that can be a complement, because we don't really want our customers to be thinking like, oh, do I use Dash or do I use Copilot? And frankly, none of them do. In a lot of ways, actually, some of the things that we do on the security front with Dash for Business are a good complement to Copilot, because as part of Dash for Business, we actually give admins, IT, like, universal visibility and control over all the different, what's being shared

• 35:55Speaker 01

  in your company across all these different platforms. And as a precondition to installing something like Copilot, or Dash, or Glean, or any of these other things, right? You know, IT wants to know, like, hey, before we, like, turn all the lights in here, like, let's do a little cleaning first before we let everybody in. And there just haven't been good tools to do that. And post-AI, you would do it completely differently. And so that's like a big, that's a cornerstone of what we do and what sets us apart from these tools. And actually, in a lot of cases, we will help those tools be adopted because we actually help them do it safely.

• 36:26Alessio Fanelli?

  Yeah. How do you think about building for AI versus people? It's like, when you mentioned cleaning up, is because maybe before you were like, well, humans can have some common sense when they look at data on what to pick versus models are just kind of like ingesting. Do you think about building products differently, knowing that a lot of the data will actually be consumed by LLMs and like agents and whatnot versus like just people?

• 36:46Speaker 01

  I think it'll always be, I aim a little bit more for like, you know, level three, level four kind of automation, because even if the LLMs like capable of completely autonomously organizing your environment, it probably would do a reasonable job. But like, I think you build bad UI when the sort of user has to fit itself to the computer versus something that you're, you know, it's like an instrument you're playing, or something where you have some kind of good partnership. And, you know, and on the other side, you don't have to do all this like manual effort. And so like, the command line was sort of subsumed by like, you know, graphical UI. We'll keep toggling back and forth. Maybe chat will be, chat will be an increasing, and especially when you bring in voice, like, will be an increasing part of the. Puzzle, but I don't think we're gonna go back to like a million command lines either.

• 37:29Speaker 01

  And then as far as like the the sort of plumbing of like, well, is this gonna be consumed by an L M or a human? Like fortunately, like you don't really have to design it that differently. I mean, you have to make sure everything's legible to the L M, but it's like quite tolerant of, you know, malformed everything. And actually, the more, the easier you make something to read for a human, the easier it is for an L M to read to some extent as well. But we really think about what's that kind of right? How do we build that right like human machine interface where you're still in control and driving? But then it's super easy to translate your intent into like the, you know, however you want your folder, setting your environment setup or like your preferences.

• 38:05Alessio Fanelli?

  What's the most underrated thing about Dropbox that maybe people don't appreciate?

• 38:09Speaker 01

  Well, I think this is just such a natural evolution for us. It's pretty true. Like when people think about the world of A I, file syncing is not like the, you know, next thing you would autocomplete mentally. And I think we also did like our first thing so well that there are a lot of benefits to that. But I think they're also, or like we hit it so hard with our first product that it was like pretty tough to come up with a sequel and we had a bit of a sophomore slump and, you know, I think actually, a lot of kids do use Dropbox through in high school or things like that, but you know, they're not, they're a lot more in the browser than their file system, right? And we know all this, but still like we're super well positioned to like help a new generation of people

• 38:52Speaker 01

  with these fundamental problems and these like that affect, you know, a billion knowledge workers around just finding, organizing, sharing your stuff and keeping it safe. And there's, there's a ton of unsolved problems in those four verbs. We've talked about search a little bit, but just even think about like a whole new generation of people like growing up without the ability to like organize their things. And yeah, search is great. And if you just have like a giant infinite pile of stuff, then search does make that more manageable. But, you know, you do lose some things that were pretty helpful in prior decades. Right? So even just the idea of persistence, stuff still being there when you come back. Like when I go to sleep and wake up, my physical papers are still on my desk. When I reboot my computer, the files are still on my hard drive.

• 39:33Speaker 01

  But then when in my browser, like if my operating system updates the wrong way and closes the browser, or if I just more commonly just declare tab bankruptcy. It's like your whole workspace just clears itself out and starts from zero. And you're like, on what planet is this a good idea? There's no concept of like, oh, here's the stuff I was working on. Yeah, let me get back to it. And so that's like a big motivation for things like Dash. Huge problems with sharing, right? If I'm remodeling my house or if I'm getting ready for a board meeting, what do I do if I have a Google Doc and an Airtable and a 10 gig 4K video? There's no collection that holds mixed format things. And so it's another kind of hidden problem hidden in plain sight. Like he's missing primitives like, yeah, files have folders, songs have playlists, links have,

• 40:18Speaker 01

  you know, there's no, somehow we miss that. And so we're building that with stacks in Dash where it's like a mixed format, smart collection that you can then, you know, share whatever you need internally, externally, and have it be like a really well-designed experience and platform agnostic and not tying you to any one ecosystem. We're super excited about that. You know, we talked a little bit about security in the modern world. Like IT signs all these compliance documents, but in reality has no way of knowing where anything is or what's being shared. It's actually better for them to not know about it than to know about it and not be able to do anything about it. And when we talked to customers, we found that there were like literally people in IT whose jobs it is to like manually go through, log into each like, log into Office, log into Workspace, log into each tool and like go comb through one by one the links that people have shared and like unshare.

• 41:03Speaker 01

  There's like an unshare guy in all these companies and that job is probably about as fun as it sounds. Like, my God. So there's, you know, fortunately, I guess what makes technology a good business is for every problem it solves, it like creates a new one.

• 41:17Alessio Fanelli?

  Right.

• 41:17Speaker 01

  So there's always like a sequel that you need. And so, you know, I think the happy version of our Act Two is kind of similar to Netflix. I look a lot of these companies that really had multiple acts and Netflix had the vision to be streaming from the beginning, but broadband and everything wasn't ready for it. So they started by mailing you DVDs, but then went to streaming and then, but the value probably the whole time was just like, let me press play on something I want to see. And they did a really good job about bringing people along from the DVD mailing off. You would think like, oh, the DVD mailing piece is like this burning platform or it's like legacy, you know, ankle weight. And they did have some false starts in that transition. But when you really think about it, they were able to take that DVD mailing audience,

• 41:58Speaker 01

  move like migrate them to streaming and actually bootstrap a, you know, take their season one people and bootstrap a victory in season two because they already had, you know, they weren't starting from scratch. And like both of those worlds were like super, it's super easy to sort of forget and be like, oh, it was all kind of destiny. But like, no, that was like an incredibly competitive environment and Netflix did a great job of like activating their act one advantages and winning an act two because of it. So I don't think people see Dropbox that way. I think people are sort of thinking about it as just in terms of our act one and they're like, yeah, Dropbox is fine. I used it 10 years ago, but like, what have they done for me lately? And I don't blame them. So fortunately we have like better and better answers to that question every year.

• 42:39Alessio Fanelli?

  And you call it like the silicon brain. So you see like dash and stacks being like the silicon brain interface basically for people?

• 42:46Speaker 01

  I mean, it's part of it. Yeah. And writ large, I mean, I think what's so exciting about AI and everybody's got their own kind of take on it. But if you like really zoom out civilisationally and like what allows humans to make progress? And you know, what, what sort of is above the fold in terms of what's really mattered? Certainly one of, I mean, there are a lot of points, but some that come to mind, like you think about things like the industrial revolution, like before that, like mechanical energy, like the only way you could get it was like by your own hands, maybe an animal, maybe some like clever sort of machines or machines made of like wood or something.

• 43:20Alessio Fanelli?

  Yeah.

• 43:21Speaker 01

  Right. But you were sort of quite like energy limited. And then suddenly, you know, the industrial revolution, things like electricity, it suddenly is like, all right, mechanical energy is now available on demand. It's like very fungible kind of, and then suddenly we consume a lot more of it and then the standard of living goes way, way, way, way up. That's been pretty limited to the physical realm. And then I believe that the large models, that's really the first time. We can kind of bottle up cognitive energy or and offload. You know, if we started by offloading a lot of our mechanic or physical busy work to machines, that freed us up to make a lot of progress in other areas. And then with AI and computing, we're like, now we can offload a lot more of our cognitive busy work to machines and then then we could create a lot more of it. Price of it goes way down.

• 44:04Speaker 01

  Importantly, like, it's not like humans never did anything physical again. It's sort of like, no, but we're more leveraged. We can move a lot more earth with a bulldozer than a shovel. And so that's like what is at the most fundamental level. What's so exciting to me about AI. And so what's the silicon brain? It's like, well, we have our human brains and then we're going to have this other like half of our brain that's sort of coming online like our silicon brain. And it's not like one or the other. They complement each other. They have very complementary strengths and weaknesses, and that's that's a good thing. There's also this weird tangent we've gone on as a species to like where knowledge work, knowledge workers have this like epidemic of of burnout, great resignation, quiet quitting. And there's a lot going on there, but I think that's one of the biggest problems we have

• 44:46Speaker 01

  is that be like people deserve like meaningful work and, you know, cancel all of it. But like in at least in knowledge work, there's a lot of own goals, you know, enforced enforced errors that we're doing where it's like, you know, on one side with brain science, like we know what makes us like productive and fortunately, it's also what makes us engaged. It's like when we can focus or when we're some kind of flow state. But then we go to work and then increasingly going to work is like going to a screen. And you're like, if you wanted to design an environment that made it impossible to ever get into a flow state or ever be able to focus, like what we have is that. And that was the thing that was just like seven, eight years ago, that blew my mind. I'm just like, I cannot understand why like knowledge work is so jacked up on this dimension.

• 45:28Speaker 01

  It's like we put ourselves in like the most cognitively polluted environment possible. And we put so much more stress on the system when we're working remotely and things like that. And, you know, all of these problems are just like going in the wrong direction. And I just, I just couldn't understand why this was like a problem that wasn't fixing itself. And like, maybe there's something Dropbox can do with this and, you know, things like DASH are the first step. But then, well, so like what? Well, I mean, now that we're like, well, why are humans in this like polluted state? It's like, well, we're just. All of the tools we have today, like this generation of tools, just passes on all of the weight, the burden to the human. Right? So it's like, here's a bajillion, 80,000 unread emails. Cool. Here's 25 unread Slack channels. Here's, we all get sort of like, it's like jittery, like, thinking about it.

• 46:13Speaker 01

  And then you look at that, you're like, wait, I'm looking at my phone, it says like 80,000 unread things. There's like no question, product question for which is the right answer. Fortunately, that's why things like our silicon brain are pretty helpful, because like, they can serve as like an attention filter. Where it's like, actually, computers have no problem reading a million things. Humans can't do that, but computers can. And to some extent, this was already happening with computing, you know, Excel is in a version of your silicon brain, or you know, you could draw the line arbitrarily. But with larger models, like now, so many of these little subtasks and tasks we do at work can be like fully automated. And I think, you know, I think it's like an important metaphor to me, because it mirrors a lot of what we saw with computing, computer architecture generally.

• 46:55Speaker 01

  It's like, we started out with a CPU, very general purpose, then GPU came along, much better at these like parallel computations. We talk a lot about like human versus machine being like substituting, or it's like, CPU, GPU, it's not like one is categorically better than the other, they're complements. Like if you have something really parallel, use a GPU, if not, use a CPU. The whole relationship, that symbiosis between CPU and GPU has obviously evolved a lot since, you know, playing Quake 2 or something. But right now, we have like the human CPU doing a lot of, you know, silicon CPU tasks. And so you really have to like redesign the work thoughtfully, such that, you know, probably not that different amounts evolved. Computer architecture, where the CPU is sort of an orchestrator of these really like heavy lifting GPU tasks.

• 47:39Speaker 01

  That dividing line does shift a little bit, you know, with every generation. And so I think we need to think about knowledge work in that context. Like, what are human brains good at? What's our silicon brain good at? Let's resegment the work. Let's offload all the stuff that can be automated. Let's go on a hunt for like anything that could save a human CPU cycle. Right. Let's give it to the silicon one. And so I think we're at the early earnings of actually being able to do something about it.

• 48:00Alessio Fanelli?

  It's funny, I gave a talk to a few government people earlier this year with a similar point, where we used to make machines to replace human labor and then the kilowatt hour was kind of like the unit of growth for a lot of countries. And now you're doing the same thing with the brain. And these data centers are kind of computational powerplants, you know, they're kind of on the band tokens. You're on the board of Meta, which is the number one donor of Lops for the open source world. The thing about open source AI is like, the model can be open source, but you need to carry a briefcase to actually maybe run a model that is not even that good compared to some of the big ones. How do you think about some of the differences in the open source ethos with like traditional software where it's like really easy to run and act on it versus like models where it's like, it might be open source, but like, I'm kind of limited, sure I can do it.

• 48:45Speaker 01

  Yeah. Well, I think with every new era of computing, there's sort of a tug of war between is this gonna be like an open one or a closed one? And, you know, there's pros and cons to both. It's not like, oh, open is always better, or open always wins. But, you know, I think you look at how the mobile, like the PC era and the internet era started out being more on the open side, like it's very modular, everybody, the sort of party that everybody could, you know, come to. Some downsides of that, security. But I think, you know, the advent of AI, I think there's a real question, like, given the capital intensity of what it takes to train these foundation models, like, are we gonna live in a world where oligopoly or cartel, there's a few companies that have the keys and we're all just like paying them rent, you know, that's one

• 49:27Speaker 01

  future. Or is it gonna be, you know, more open and accessible? And I'm like, super happy with how that's, just I find it exciting on many levels with all the different hats I wear about it. Fortunately, you've seen in real life, yeah, even if people aren't, you know, bringing GPUs on a plane or something, you've seen like the price performance of these models improve 10 or 100x year over year, which is, it's sort of like many Moore's Laws compounded together for a bunch of reasons. Like, that wouldn't have happened without open source, right? You know, for a lot of same reasons, it's probably better that we can, anyone can sort of spin up a website without having to buy an internet information server license, like there was some alternative futures, like things are Linux and really good. And there was a good balance of trade too, where

• 50:10Speaker 01

  like people contribute their code and then also benefit from the community. Returning the favor, I mean, you're seeing that with open source. So you wouldn't see all this like, you know, this flourishing of research and of just sort of the democratization of access to compute without open source. And so I think it's been like phenomenally successful in terms of just moving the ball forward and pretty much anything you care about, I believe, even like safety.

• 50:34Alessio Fanelli?

  Yeah.

• 50:34Speaker 01

  If you can have a lot more eyes on it and transparency instead of just something is happening in those three places with nuclear power plants attached to them, right? So I think it's been awesome to see. And then, and again, for like wearing my Dropbox hat, like anybody who's like scaling a service to millions of people, again, are probably not using like frontier models for every request. It's, you know, there are a lot of different configurations, mostly of smaller models. And even talk about getting on the device, like, you know, you need this whole kind of constellation of different options. So open source has been great for that.

• 51:06Alessio Fanelli?

  And you were one of the first companies in the cloud repatriation. You kind of brought back all the storage into your own data centers. Where are we in the AI way for that? I don't think people really care today to bring the models in house.

• 51:19Speaker 01

  Yeah.

• 51:19Alessio Fanelli?

  Do you think people will care in the future? Like especially as you have more small models, like you want to control more of the economics? Or are the tokens so subsidized that like it just doesn't matter? It's more like a principle standard?

• 51:29Speaker 01

  Yeah. Yeah. I mean, I think this is another one where like thinking about the future is a lot easier or if you start with the past. So, I mean, there's definitely this like big surge in demand as like there's sort of this FOMO driven bubble of like all of big tech taking their own things and shipping them to Jensen for a couple of years. And then you're like, all right, well, first of all, we've seen this kind of thing before and in late 90s with like fiber, you know, this huge race to like own the internet, the only information superhighway, literally, and then way overbuilt. And then there was this like crash. I don't know to what extent, like maybe it is really different this time or, you know, maybe if we create AGI that will sort of solve the rest of it or we'll just have a different set of things to worry about.

• 52:12Speaker 01

  But, you know, the simplest way I think about it is like this is sort of a rent not buy phase. Because I wouldn't want to be, we're still so early in the maturity. I wouldn't want to be buying pallets of 286s at a 5X markup when the 386 and 486 and Pentium and everything are clearly coming, they're around the corner. And again, because of open source, there's just been a lot more competition at every layer in the stack. And so product developers are basically beneficiaries of that. The things we can do, the sort of cost decents I was looking at a year or two ago to provide different capabilities in the product, you know, cut, you know, slash them by 10, 100, 1,000 X. I think about coming back around, I think at some point you have to believe that the

• 52:56Speaker 01

  sort of supply and demand will even out as it always does. And then there's also like non-Nvidia stacks, like the Grok or Cerebras or some of these custom silicon companies are super interesting and outperform Nvidia stack in terms of latency and things like that. So I guess it's a pretty exciting change. I think we're not close to the point where we were with hard drives or storage when we sort of went back from the public cloud. Because there it was like, yeah, the cost curves are super predictable. We know what the cost of a hard drive and a server and terabyte of bandwidth and all the inputs are going to just keep going down, riding down this cost curve. But to rely on the public cloud to pass that along is sort of, we need a better strategy than relying on the kindness of strangers. So we decided to bring that in-house and still do, and we still get a lot of advantages.

• 53:41Speaker 01

  That said, the public cloud has scaled and been a lot more reliable and just good all around than we would have predicted. Because actually, back then, we were worried, is the public cloud going to even scale fast enough to keep up with us? But yeah, I think we're in the early innings. It's a little too chaotic right now. So I think renting and preserving agility is pretty important in times like these.

• 54:01Alessio Fanelli?

  Yeah, we just went to the Cerebras factory to do an episode there. We saw one of their data centers inside. Yeah, it's kind of like, okay, if this really works.

• 54:09Speaker 01

  It can be a lot of influence.

• 54:10Alessio Fanelli?

  Yeah, exactly. It kind of changes everything.

• 54:13Speaker 01

  And that is one of those things, like this is one where you could just have these things that just like, okay, there's just like a new kind of piece on the chessboard, like recalc everything. So I think there's still, I mean, that's like not that likely, but I think this is an area where it actually could, you could have these sort of like, and out of nowhere, suddenly, all of a sudden, you know, everything's different.

• 54:32Alessio Fanelli?

  Yeah. I know one of the management books he references, Andy Grove's, How Many the Paranoid Survived?

• 54:37Speaker 01

  Yep.

• 54:37Alessio Fanelli?

  Maybe if you look at Intel, they did a great job of memory to chip, but then it's like maybe CPU to GPU, they kind of missed, missed that thing.

• 54:44Speaker 01

  Yeah.

• 54:45Alessio Fanelli?

  How do you think about staying relevant for so long? It's been 17 years, you've been doing Dropbox. What's the secret? And maybe we can touch on founder mode and all of that.

• 54:54Speaker 01

  Yeah. Well, first, what makes tech exciting and also makes it hard is like, there's no standing still, right? And your customers never are like, oh no, we're good now. They always want more or just, and then the ground is shifting under you, or it's like, oh yeah, well, files are not even that relevant to the modern, I mean, it's still important, but like, you know, so much is tilted elsewhere. So I think you have to like always be moving and think about, on the one level, like what is, and think of these different layers of abstraction, like, well, yeah, the technical service we provide is file syncing and storage in the past, but in the future, it's going to be different. The way Netflix had to look at, well, technically, we mail people physical DVDs and fulfillment centers, and then we have to switch like streaming and codecs and bandwidth and data centers. So you do have to think about that level. But then it's like, what's the evergreen problem we're

• 55:39Speaker 01

  solving is an important problem. Can we build the best product? Can we get distribution? Can we get a business model? Can we defend ourselves when we get copied? And then having like some context of like, history has always been like one of them, reading about the history, not just in tech, but of business, or government, or sports, or military. These things that seem like totally new, you know, and to me would have been like totally new as a 25 year old, like, oh my God, the world's completely different. Everything's going to change. You're like, well, there's not a lot of great things about getting older, but you do see like, well, no, this actually has like a million like precedents. And you can actually learn a lot from, you know, about like the future of GPUs from like, I don't know, how, you know, how Formula One teams work. Or you can draw all these like weird analogies that are super helpful and guiding you from first principles, through

• 56:24Speaker 01

  a combination of first principles and like past context. But like, you know, build shit we're really proud of. Like, that's a pretty important first step. And really think about, like, you sort of become blind to like how technology works. That's just the way it works. And you know, something like carrying a thumb drive, you're like, well, I'd much rather have a thumb drive than like literally not have my stuff. Or like have to carry a big external hard drive around. So you're always thinking like, oh, this is awesome. Like I ripped CDs and these like MP3s and these files and folders. This is the best. But then you miss, on the other side, you're like, this isn't the end, right? MP3s and folders is like, and Apple comes along, it's like, this is dumb. You should have like a catalog, artist playlist, you know? Then Spotify is like, hey, this is dumb. Like you should, why are you buying these things à la carte? It's the internet, you should have access to everything. And then by the way, why is this like such a single player experience?

• 57:07Speaker 01

  You should be able to share and they should have, they should be AI curated, et cetera, et cetera. And then a lot of it is also just like drawing, connecting dots between different disciplines, right? So a lot of what we did to make Dropbox successful is like, we took a lot of the consumer playbook, applied it to business software from a virality and kind of ease of use standpoint. And then, you know, I think there's a lot of you can draw from the consumer realm and what's worked there. And that hasn't been ported over to business, right? So a lot of what we think about is like, yeah, when you sign into Netflix or Spotify or YouTube or any consumer experience, like what do you see? Well, you don't see like a bunch of titles starting with AA, right? You see like this whole, and it went on evolution, right?

• 57:48Speaker 01

  Like we talked about music and TV went through the same thing, like 10 channels over the air broadcast to 30 channels, 100 channels. But then some of like 1,000 channels, you're like, this has totally lost the plot. So we're sort of in the 1,000 channels era of productivity tools where we're just like, wait, wait, we just need to rethink the system here. And we don't need another 1,000 channels. We need to redesign the whole experience. And so, and I think the consumer experiences that are like smart, you know, when you sign into Netflix, it's not like 1,000 channels. It's like here are a bunch of smart defaults, even if you're a new signup, we don't know anything about you. But because of what the world is watching, here are some, you know, reasonable suggestions. And then it's like, okay, I watched Drive to Survive, I didn't watch Squid Game. You know, the next time I sign in, it's like, it's a learning system, right?

• 58:29Speaker 01

  So a combination of design, machine learning, and just like the courage to like rethink the whole thing. I think that's, that's a pretty reliable recipe. And then you think, you're like, all right, there's all that intelligence in the consumer experience. There's no filing things away, everything is sort of auto curated. For you and sort of self optimizing, then you go to work and you're like, there's not even an attempt to incorporate any intelligence or organization anywhere in this experience. And so like, okay, can we do something about that?

• 58:57Alessio Fanelli?

  You know, you're one of the last founder CEOs, like you would talk, then you like Toby Lute, some of these folks. How, how does that change?

• 59:05Speaker 01

  I'm like 300 years old and why?

• 59:06Alessio Fanelli?

  No, I would say like when you run, when you run a company, like you've had multiple executives over the years, like, how important is that for the founder to be CEO and just say, hey, look, we're changing the way the company and the strategy works. It's like, we're really taking this seriously versus like, you could be a public CEO and be like, hey, I got my earnings call and like, whatever, I just need to focus on getting the, the right numbers. Like, how does that change the culture in the company?

• 59:28Speaker 01

  Yeah, well, I think it's sort of dovetails with the founder mode whole thing. You know, I think founder mode is kind of this Rorschach test. It's sort of like ill specified, so it's sort of like whatever you, you know, it is whatever you see it. I think it's also like a destination you get to more than like a state of mind. Right? So if you think about, you know, imagine someone, there was something called surgeon mode, you know, giving a med student a scalpel on day one, it's like, okay, hold up. You know, so there's something to be said for like experience and conviction and you know, you're going to do a lot better. A lot of things are a lot easier for me, like 17 years into it than they were one year into it. I think part of why founder mode is so resonant is, or it's like striking such a chord with

• 1:00:11Speaker 01

  so many people is, yeah, there's, there's a real power when you have like a directive, intuitive leader who can like decisively take the company like into the future. It's like, how the hell do you get that? And I think every founder who makes it this long, like, kind of can't help it but to learn a lot during that period. And you talk about the, you know, Steve Jobs or Elons of the world, they, they did go through like wandering, a period of like wandering in the desert or like nothing was working and they weren't the cool kids. I think you either sort of like unsubscribe or kind of get off the train during that. And I don't blame anyone for doing that, that many times where I thought about that. But I think at some point you sort of, it all comes together and you sort of start being able to see the matrix. So you've sort of seen enough and learned enough. And as long as you keep your learning rate up, you can kind of surprise yourself in terms

• 1:00:55Speaker 01

  of like how capable you can become over a long period. And so I think there's a lot of like founder CEO journey, especially as an engineer, like You know, I never like set out to be a CEO. In fact, like, the more I like understood in the early days what CEOs did, the more convinced I was that I was like not the right person, actually. And it was only after some like shoving by a previous mentor, like, "Don't, just go try it, and if you don't like it, then you don't have to do it forever." So I think you start founder mode, you're sort of default that because there's like, you realize pretty quickly, like, nothing gets done in this company unless the founders are literally doing it by hand. Then you scale and then you're like, you get, you know, a lot of actually pretty good advice that like, you can't do everything yourself, like, you actually do need to hire people

• 1:01:38Speaker 01

  and like give them real responsibilities and empower people. And that's like a whole discipline called like management that, you know, we're not figuring out for the first time here. But then you, then there's a tendency to like lean too far back. You know, it's tough, and if you're like a 30-year-old and you hire a 45-year-old exec from, you know, a high-flying company, and a guy who's running like a $10 billion P&L and came to work for Dropbox where we were like a fraction of a billion-dollar P&L, and what am I gonna tell him about sales, right? And so you sort of recognize pretty, like, "I actually don't know a lot about all these different disciplines, and like, maybe I should lean back and like let people do their thing." But then you can create this like, if you lean too far back out, you create this sort of like vacuum, leadership vacuum, where people are like, "What are we doing?"

• 1:02:23Speaker 01

  And then, you know, the system kind of like, nature reports a vacuum, it builds all these like kind of weird structures just to keep the thing like standing up. And then at some point, you learn enough of this, that you're like, "Wait, this is not how, this should be designed." And you actually get like the conviction and the, you learn enough to like know what to do and things like that. And then on the other side, you lean way back in. I think it's more of like a table flipping, where you're like, "Hey, this company is like not running the way I want it." Like something, "I don't know what happened, but it's gonna be like this now." And I think that that's like an important developmental stage for a founder CEO. And if you can do it right and like make it to that point, like, then the job becomes like a lot of fun and exciting and good things happen for the company, good things happen for your customers.

• 1:03:06Speaker 01

  But it's not, it's like a really rough, you know, learning journey.

• 1:03:09Alessio Fanelli?

  It is, it is. I've had many therapy sessions with founders, CEOs. Let's go back to the beginning. Like today, the AI wave is like so big that like a lot of people are kind of scared to jump in the water. And when you started Dropbox, one article said, fortunately, the Dropbox founders are too stupid to know everyone's already tried this. In AI now, it kind of feels the same. You have a lot of companies that sound the same, but like, none of them are really working. So obviously, the problem is not solved. Do you have any advice for founders trying to navigate like the D M A's today? And like what they should do, what are like counterintuitive things maybe to try?

• 1:03:45Speaker 01

  Well, I think like, you know, bringing you guys some of what we've covered, I think there's a lot of very common kind of category errors that founders make. One is, you know, thinking of starting from the technology versus starting from like a customer, starting from a use case. And I think every founder has to start with what you know, like you're, yeah, you know, maybe if you're an engineer, you know how to build a product, but don't know any of the other next, you know, hurdle, you don't know much about the next hurdles you have to go through. So I think, I think the biggest lesson would be you have to keep your personal growth curve out of the company's growth curve. And for me, that meant you have to be like super systematic about training up what you don't know. Cause no one's gonna do that for you. Your investors aren't gonna do that. Like literally no one else will do that for you.

• 1:04:26Speaker 01

  And so then it, then you have to have like, all right, well, and I think the most important, one of the most helpful questions to ask there is like, in five years from now, what do I wish I had been learning today? In three years from now, what do I wish I had been in one year? You know, how will my job be different? How do I work back from that? And so for example, you know, when I was just starting in 2007, it really was just like coding and talking to customers. And it's sort of like the Y C ethos, you know, makes so many people want and coding and talking to customers are really all you should be doing in that early phase. But then if I were like, all right, what that's sort of Y C phase, what's, what are the next hurdles? Well, a year from now, then I'm gonna need, but to get people, we're gonna need fundraised, like raise money. Okay, to raise money, we're gonna have to answer all these questions. We have to, blah, blah. So you work back from that and you're like, alright, we need to become like an expert in like venture capital financing and then the circle keeps expanding.

• 1:05:11Speaker 01

  Then if we have a bunch of money, we're gonna need accountants and lawyers and employees and I'm gonna have to start managing people. Then two years would be like, well, we're gonna have this like product, but then we're gonna need users, we're gonna need money, revenue. And then in five years, it'd be like, yeah, we're gonna be like tangling with like Microsoft, Google, Apple, Facebook, everybody. And like somehow, we're gonna have to like deal with that. And then that's like what the company's got to deal with. And as CEO, I'm gonna be responsible for all that. But then like my personal growth, there's all these skills I'm gonna need. I'm gonna need to know like what marketing is and like what finance is and how to manage people, how to be a leader, whatever that is. And so, and then I think one thing people often do is like, oof, like that, it's like imposter syndrome kind of stuff. You're like, oh, it seems so remote or far away that, or I'm not comfortable speaking

• 1:05:55Speaker 01

  publicly, or I've never managed people before, I haven't this, I haven't that. And even learning a little bit about it makes it feel even worse. You're like, now, I thought I didn't know a lot, now I know I don't know a lot. Right. Part of it is more technical, like how do I learn all these different disciplines and sort of train myself? And a lot of that's like reading, you know, having founders or community that are sort of going through the same things. That was how I learned. Maybe reading was the single most helpful thing, more than any one person, or talking to people, like reading books. But then there's a whole mindset piece of it, which is sort of like, you have to cut yourself a little bit of slack. Like, you know, I wish someone had sort of sat me down and told me like, dude, you may be an engineer, but like look all the tech founders that, you know, tech CEOs that you admire, like they actually all, you know, almost all of them started out as engineers. They learned the business stuff on the job.

• 1:06:40Speaker 01

  So like this is actually something that's normal and achievable, you're not like broken for not knowing. No, those people didn't, weren't like, didn't come out of the womb with like shiny hair and a Armani suit. You know, you can learn this stuff. So even just like knowing it's learnable. And then second, like, but I think there's a big piece of it around like discomfort, where it's like, oh, I mean, we're like kind of pushing the edges. I don't know if I want to be CEO, or I don't know if I'm ready for this, this, this. Like, learning to like walk towards that when you want to run away from it. And then lastly, I think, you know, just recognizing the time constant. So five weeks, you're not gonna be a great leader, manager, or great public speaker, or whatever, you know, think any more than you'll be a great guitar player, or play support that well, or be a surgeon. But in like five years, like actually you can be pretty good at any of those things.

• 1:07:23Speaker 01

  Maybe you won't be like fully expert, but you have like a lot more latent potential. People have a lot more latent potential than they fully appreciate. But it doesn't happen by itself. You have to carve out time and really be systematic about unlocking it.

• 1:07:36Alessio Fanelli?

  How do you think about that for building your team? I know you're a big Pat's fan.

• 1:07:39Speaker 01

  Yes.

• 1:07:40Alessio Fanelli?

  That's a great example of building a dynasty on like some building blocks and bringing people.

• 1:07:44Speaker 01

  Totally.

• 1:07:45Alessio Fanelli?

  When you're building a company, like how much slack do you have people on? Hey, you're gonna learn this versus like, how do you measure like the learning rate?

• 1:07:52Speaker 01

  Yeah.

• 1:07:53Alessio Fanelli?

  The people you hire and like, how do you think about picking and choosing?

• 1:07:55Speaker 01

  Great question. It's hard. What you want is a balance, right? And we've had a lot of success with great leaders who actually grew up with a company. Started as an IC engineer or something, then made their way to whatever level. Our exec team is populated with a lot of those folks. But there's also a lot of benefit to experience and having seen different environments and kind of been there, done that. And there's a lot of drawbacks to kind of learning by trial and error only. And then even your high potential people like can go up the learning curve faster if they have like someone experienced to learn from. Now, like experience isn't a fantasy. Either like you can have various organ rejection or misfit or like overfitting from their past experience or cultural mismatches or you name it.

• 1:08:36Speaker 01

  I've seen it all. I've kind of gotten all the mistake merit badges on that. But I think it's like constructing a team where there's a good balance. Like, okay, for the high potential folks who are sort of in the biggest jobs of their lives, can, do they either have someone that they're, is managing them, that they can learn from? You know, as a CEO, part of your job or as a manager, like you have to like surround or help support them. So getting them mentors or getting first time execs like mentors who have been there, done that, or getting them in like, you know, there's usually for any function, there's usually like a social group, like, oh, chiefs of staff of Silicon Valley. Okay. Like, you know, there's usually these informal kind of communities you can join. And then, yeah, you just don't want to be too rotated in one direction or the other. Because we've done it, we've like overdone it on the high potential piece,

• 1:09:19Speaker 01

  but then like everybody's kind of making dumb mistakes.

• 1:09:22Alessio Fanelli?

  Right, yeah, yeah, yeah.

• 1:09:23Speaker 01

  The bad mistakes are the ones where like either you're making it multiple times or like these are known, knowns to the industry. But if they're not known, if they're like unknown unknowns to your team, then you're you have a problem. And then again, if you have too much, if you've just only hire external people like then you're sort of at the mercy. You'll be like whatever random average of whatever culture or practices they bring in can create resentment or like lack of career opportunities. So it's really about how do you get, you know, it doesn't really matter if it's like exactly fifty fifty. I don't think about a sort of perfect balance, but you just need to be sort of tending that garden continuously.

• 1:09:57Alessio Fanelli?

  Also, Drew, just to wrap, do you have any call to actions, like who should come work at Dropbox? Like who should use Dropbox, anything you want, you want to tell people?

• 1:10:06Speaker 01

  Well, I'm super, I mean, today is a super exciting day because we just launched Dash for Business. And, you know, we talked a little bit about the product. It's like universal search, universal access control, a lot of rethinking, sharing for the modern environment. But, you know, what's personally exciting, you know, could talk about the product, but like the, it's just really exciting for me to like, yeah, this is like the first, like most major and most public step we've taken from our kind of Dropbox one point O roots. And there's probably a lot of people out there who either like grew up not using Dropbox or like, yeah, I used Dropbox like ten years ago and it was cool, but I don't do that much of fun. So I think there's a lot of new reasons to kind of tune into what we're doing.

• 1:10:44Speaker 01

  And it's a lot of, it's been a lot of fun to, I think, like the sort of the AI era has created all these new like paths forward for Dropbox that wouldn't have been here five years ago. And then, yeah, to the founders, like, you know, hang in there, do some reading.

• 1:11:00Speaker 01

  And don't be too stressed about it. So we're pretty lucky to get to do what we do.

• 1:11:04Alessio Fanelli?

  Yeah, watch the Pat's Doug battery on Apple TV.

• 1:11:08Speaker 01

  Yeah, Bill Belichick, I'm still Pat's fan, really got an F1. So we're technology partners, McLaren, they're doing.

• 1:11:14Alessio Fanelli?

  Were you a McLaren fan before you were technology partner? So did you become partners because?

• 1:11:19Speaker 01

  It's sort of like co-evolved, yeah. I mean, I was a fan beforehand, but I'm like a lot more of a fan now, as you'd imagine.

• 1:11:24Alessio Fanelli?

  Awesome. Well, thank you so much for the time, Drew. This was great.

• 1:11:27Speaker 01

  Yeah, it's a lot of fun.

• 1:11:28Alessio Fanelli?

  Yeah.

• 1:11:28Speaker 01

  Thanks for having me.