Build a workspace of AI agents

How can every single person build a personal AI protégé and then accumulate (and share) a host of other assistants? In this episode, we dive into the world of no-code AI with Scott Meyer from Chipp.ai. We discuss AI tooling for people that can't code, the cultural shift that needs to happen for widespread AI adoption in businesses, and the predicted growth trajectory of AI assistant that you can own. Featuring: Chris Benson – Website https://chrisbenson.com, GitHub https://github.com/chrisbenson, LinkedIn https://www.linkedin.com/in/chrisbenson, X https://x.com/chrisbensonDaniel Whitenack – Website https://www.datadan.io/, GitHub https://github.com/dwhitena, X https://x.com/dwhitenaScott Meyer - LinkedIn https://www.linkedin.com/in/scottdavidmeyer, X https://x.com/MrScottMeyerSponsors: Domo https://ai.domo.com – The AI and data products platform. Strengthen your entire data journey with Domo’s AI and data products. Show Notes: Chipp.ai https://www.chipp.ai/Chipp.ai's Discord https://discord.com/invite/TMfcNqFWSt

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

• 0:03Speaker 00

  Welcome to Practical AI, the podcast that makes artificial intelligence practical, productive, and accessible to all. If you like this show, you will love The Change Log. It's news on Mondays, deep, technical interviews on Wednesdays, and on Fridays, an awesome talk show for your weekend enjoyment. Find us by searching for The Change Log wherever you get your podcasts. Thanks to our partners at fly.io. Launch your AI apps in five minutes or less. Learn how at fly.io.

• 0:44Daniel Whitenack?

  Welcome to another episode of the Practical AI Podcast. This is Daniel Whitenack. I'm CEO at Prediction Guard, and I'm joined as always by my co-host, Chris Benson, who is a Principal AI Research Engineer at Lockheed Martin. How are you doing, Chris?

• 1:00Speaker 02

  Oh, I'm feeling pretty chipper today. It's a, it's a good day to talk about AI.

• 1:05Daniel Whitenack?

  Yeah, yeah, I feel, feel quite, quite chipper as well, especially as we've got our guest today, Scott Meyer with us, who's founder and CEO at Chip, which is, you can find at chipp.ai, I believe, is, is the link. But yeah, Chip is awesome. Also, Scott is awesome. And also Scott is, is along a good friend because he's a fellow member of the Silicon Prairie, not living on the, on the coast, but out here in the middle somewhere where AI is really blossoming, if you didn't know.

• 1:40Speaker 02

  It is. And it gives an unfair advantage for those of us in non-metro areas, you know, like the ability to leverage AI to have the power of 10 people in a place that doesn't have enough people to do the job. It's perfect. It's a, it's a perfect solution. So it's great to be here live from Fargo, just like the movie. It's fantastic to see you all and be heard by all of you listening.

• 2:01Daniel Whitenack?

  Yeah, yeah, Scott, well, we'll get into all the cool stuff you know you're doing with Chip and some of the things you've learned through that, but I'm wondering if you know you work in the space of, I guess we might put it like low code, no code, AI assistant builders. So for maybe audience members that aren't as familiar with that space, or maybe they're just kind of wondering what's out there, you know, as of, as of today, could you paint a little bit of a picture for us for kind of what sorts of tools are out there? And then maybe that would kind of motivate some of the unique things that you thought

• 2:43Daniel Whitenack?

  should be out there, but weren't, which would maybe kind of highlight some of the things you're doing with Chip.

• 2:48Speaker 02

  Yeah, no, it's great to be here. I think the stat that blows my mind is that almost 50% of Americans use AI every week, but 7% of businesses use AI, which is obviously a lie, because 50% of Americans are using AI every week and they work at those companies. What's happening is the businesses aren't, they have no idea what's going on. It's like the early days of cell phones when everyone would come to work with their own cellphone, their own laptop, do whatever they wanted to. And eventually we got to this point where you get a company email, you get company apps, you get like the standard way to do it. I think the risk right now is that, and the opportunity, is those who are willing to have agency and try stuff have unfair advantage, right? So I can go do my work with AI and if my colleagues don't know, and I don't have a

• 3:31Speaker 02

  culture of sharing, like all of a sudden I'm a super, superhuman. The number one thing I tell businesses when I meet with them is you should have a lunch and learn once a month and just have people say what they're doing. Because just that horizontal sharing of AI practices and ideas is all you need to build a culture of acceptance. And what makes AI so unique is it's not top down. It's not the CIO or CTO saying, I bought this thing, you guys all go use it. It's each individual figuring out how they can use it for their specific tasks. And what I've seen is admin assistants, you know, marketers, interns, right? They're all going to use it differently and often even know better how to use it because they're the ones doing the tasks. And that kind of motivated what we built with Chip, which is how do we just make AI as

• 4:13Speaker 02

  easy as possible to use? Our, you know, kind of our motto is AI for all. And I think I've spent most of my professional career working on bridging a digital divide because maybe like you, you know, people that work and live alongside me in Fargo aren't always taking advantage of the latest technology, right? And so I kind of feel like it's both a passion and mission to bring what's happening and make it accessible to those around me. In 2009, I started my first company and I was trying to tell businesses, there's this thing called social media they should use, right? Before, there were Facebook pages and Facebook ads. And it feels like that to me again, almost 20 years later, where it's like, this amazing power is right here and the best time to start learning is now. And with tools

• 4:54Speaker 02

  like Chip and others that we can talk about, it's actually better now than ever for people who aren't technical because it's not about technical ability, it's about knowledge and agency. And I think we all have that. So happy to give a landscape. I think that already went off track from your question, but hopefully that gives you the starting point.

• 5:11Daniel Whitenack?

  No, it's awesome. What would you say are kind of some of those things that might make AI hard to use? And here, you know, mostly we're talking, of course, we've talked about a lot of things in the show, but mostly we're talking about kind of what typical people would consider AI now, which would be kind of generative AI, language models, maybe vision models, et cetera. So like, what can make those difficult to use or how might people get disillusioned as they're exploring the technology?

• 5:43Speaker 02

  I'll say almost, almost every excuse people have not to use AI tools is fear. They are scared of a blank page. And this is the same with technology for 20 years. I taught entrepreneurship, I started entrepreneurship centers and all these students with amazing ideas. And you know what? 90% of them didn't do anything because they had to actually go do something, right? And it's like, You just have to start and I'm convinced the biggest challenge in AI is change management. It's just getting people to to start. And I think this happened when Google first came out, you know, it's a blank screen, blank prompt window. Like, what do I say when I can say anything? It's actually quite intimidating. And so that's the challenge, I think, with AI is like, anything's possible.

• 6:28Speaker 02

  So where do you start? I tell everybody the best place to start is to create your digital protege. Like, just tell AI what you do and have it help you do those things. AI is great at what you hate. And so find those things that you hate doing or that take a lot of time and start there. You've maybe seen that quote, I really love that, you know, I want AI to do my dishes and laundry. So I can do more art and music, not AI to do art and music. So I can do more dishes and laundry. Right? So I think we all have dishes and laundry in our day to day life. And so let's use AI there first, because that'll be the, you'll get more motivated to do fewer financial analyses or fewer, I don't know, copy editing, because that's kind of annoying than you would like making music. Because maybe that's fun for you, right?

• 7:12Speaker 02

  So start with things that you don't like. One thing I find fascinating about research on AI is actually having knowledge makes you better positioned to use AI. I think about AI as like the rebirth of the Renaissance person. It's like, if I want to create a picture on AI that looks like Picasso, but I don't know Picasso's name, it's really hard to describe that, right? If I want to make a blueprint of a Georgian architecture building, like, how do I explain that? If I don't know what Georgian architecture is. And so whatever area you live in or work in or care about, you have like expertise, right? You can talk about it all day. And that's a great place to start with AI, because you can go say those words, like, give me, I don't know,

• 7:55Speaker 02

  a hierarchy of Pokemon characters and you can name all the things and have it rank order it. Like, I have no idea what I would say for that, right? But I can talk all day about saunas and have the AI help me improve my sauna, find new water buckets, look at, you know, different ratios of time in the sauna, like, because I care about that. So find some things that you know about, that you're passionate about and start asking AI about it so you can go deeper. I love it.

• 8:17Chris Benson?

  I'm curious, a quick follow up on that, you know, because you raised a point that I hadn't really thought about, but I've observed it many, many times and you've kind of, you brought it to the surface here with, I see people who are totally comfortable getting on the search engine of their choice and searching topics and they've been doing that for years. But as soon as they pull up, you know, a chat with a given model, they're really struggling with that and they're really, that's what, you know, like from a, I'm just curious as you've, have you clearly thought about this quite a lot, what is the difference and why are people so easy to go to search and yet struggling with, with that model, you know, that has the same text box in front of it?

• 9:00Speaker 02

  Part of it's exposure, right? Just history, but I also think there's something quite vulnerable about AI where it's really a two-way conversation. Search engine is, you know, very much like, like the old card catalogs, you know, I remember my first year of elementary school, I learned card catalog and then the next year was told never have to touch that again. But it's the same, that works the same, right? I'm just going to go find something. But with AI, it's probing back and forth and actually you can get, you can get pushback and it kind of identifies how you're thinking about things. So I think there's some vulnerability around that and plenty of like blank page problem of just not knowing where to start. So start by creating a protege, start by diving into areas you care about. And I always tell people a great framework to get started is what I call the Ripe framework. So R-I-P-E.

• 9:44Speaker 02

  And it's just a way of like four sentences to put into AI to get good answers, which is the role. So like you are an expert, I don't know, copy editor, the instruction, like read through my paper and improve it, parameters. So make sure it's very concise and don't repeat a lot of the same points and examples, like here's a paper I wrote before that shows my kind of tone. You know, if you just do those four things, a role, instruction, parameter, example, like you're going to get awesome output that's personalized and much more effective and less robotic than just going there and saying, write me a paper.

• 10:17Daniel Whitenack?

  Yeah, I've had this kind of hypothesis, I guess, going around in my mind. I'm curious, Scott, on your, on your take on this because you've seen a lot of people now, you're always interacting with people on Discord or wherever, you know, trying to get their assistance to do this or that. What have you found to be kind of the qualities that make up someone who, who is just really proficient at kind of honing in the, the instructions, the data integration, the, the configuration of AI systems? My hypothesis is sort of this is almost like a,

• 10:59Daniel Whitenack?

  I, I think if we took a bunch of hostage negotiators and had them log into, to AI systems to try to, you know, either get them to do things that they wanted them to do or to jailbreak them, I think they would be like amazing at this because a lot of times it seems to me, you know, not that I feel in danger physically or something, but it's like people can get disillusioned with this. It's like not quite what I want. How do I get you to do what I want you to do? How do I like warm you up to this idea? So yeah, I'm curious on the qualities that you've seen in terms of people that have become good at configuring these

• 11:41Daniel Whitenack?

  systems, prompting, understanding how to, you know, pull in integrations or when and where to do that. Any thoughts?

• 11:49Speaker 02

  Yeah. I mean, people who are great at this are kindergarten teachers or parents of three-year-olds.

• 11:56Daniel Whitenack?

  Maybe also hostage negotiators.

• 11:58Speaker 02

  Also, it's basically the same job title.

• 12:01Daniel Whitenack?

  There's some similarity there, maybe?

• 12:03Speaker 02

  Yeah. I mean, think about talking, I mean, people I say an intern, but that's even too, too experienced. Think about talking to my three-year-old Sebastian. If I tell him three things to go do in order, there's no way he's gonna get all three of them done, right? Like, go to the bathroom, pick out some shoes, grab your snack, go to the car. Like, that's not happening. I have to be like, go to the bathroom. Good, now this, right? And now this. It's very step by step. And I think what's interesting is there's two models or two types of models emerging in AI. And you guys maybe have your own language for this. But, you know, I think about linear models like 4.0, Claude Sonnet 3.5, and we have reasoning models now like 0.3 DeepSeek and now Sonnet 3.7. And it's like, the reasoning models actually,

• 12:48Speaker 02

  that's like talking to an intern who you can give a ton of stuff and you just let it go. But if you're doing a linear model, that's very much need to do that step by step. First do this, then do this, then do this. Because the biggest, I think, frustration people have is that AI too quickly tries to get to an answer before it has all the details and things get lost. And so with chip, you know, you can prompt your AI tool and then anyone can use it. And so what we've found is like flipping the relationship is really powerful where the AI prompts you to get what it needs and then gives an answer. So you can even, you know, on chip, you can build this in so you don't have to type it every time. But on any AI tool, you might say like, before you write the paper, before you create the, you know, strategy, before you create the, I don't know,

• 13:32Speaker 02

  the press release, make sure to ask me these three things, right? And force it to get all of that information step by step, just like you do with a three-year-old. And then you go to school and then you write the paper and then you do the thing, right? So I think that's really fascinating though, seeing that divergence with reasoning, which is like, don't go step by step, just give all the context and it's gonna work through it on its own versus the three-year-old linear that's like, needs that guidance. So, so yeah, I think, and the end of the day, hostage negotiator and parents, you got this.

• 14:16Speaker 04

  Well, friends, today's ever-changing AI landscape means your data demands more than the narrow applications and single-model solutions that most companies offer. Domo's AI and data products platform is a more robust all-in-one solution for your data. It's not just ambitious, it's practical and adaptable, so your business can meet those new challenges with ease. With Domo, you and your team can channel AI and data into innovative uses that deliver measurable impact, and their all-in-one platform brings you trustworthy AI results without having to overhaul your entire data infrastructure,

• 14:57Speaker 04

  secure AI agents that connect, prepare, and automate your workflows, helping you and your team to gain insights, receive alerts, and act with ease through guided apps tailored to your role, and the flexibility to choose which AI models you want to use. Domo goes beyond productivity, it's designed to transform your processes, helping you make smarter and faster decisions that drive real growth, all powered by Domo's trust, flexibility, and their years of expertise in data and AI innovation. Data is hard, Domo is easy. Make smarter decisions and unlock your data's full potential with Domo. Learn more today

• 15:38Speaker 04

  at ai.domo.com. Again, that's ai.domo.com.

• 15:50Daniel Whitenack?

  So Scott, maybe we'll come back to kind of the tooling itself. Could you maybe kind of circle back and describe some of the... Maybe people aren't familiar with some of the kinds of tools that are out there, especially, you know, maybe there's programmers that have interacted with APIs that are listening to the show. Maybe there are people that have explored one tool or another. Maybe there's people that haven't explored anything yet. So could you maybe just help us kind of form a mental model for the kinds of AI tools that are out there? And then maybe that would lead into a discussion about kind of some of the things that that were really on your mind

• 16:32Daniel Whitenack?

  in terms of needs that weren't being addressed in that ecosystem.

• 16:37Speaker 02

  Yeah, I mean, if you if you want to think of like a simple two by two matrix, I think there's a really clear like vertical versus horizontal and like closed versus open dichotomy. So you can think about horizontal tools doing a lot of things across modes, right? So ChatGPT can write, it can create images, it can code. It's really good at all of those, but if you want to just make images like Midjourney is probably better, right? It's a vertical image generation tool or Pika is really good at video generation, which some of the general horizontal tools aren't as good at. And, you know, my sense is like horizontal is going to win, but there's always going to be a need for people who want the Maserati of AI, right?

• 17:18Speaker 02

  If you're only doing code, like you're going to probably be in cursor going deep into like using these tools. Whereas someone like me, I'm going to do the vibe coding where I can use a tool like lovable or Bolt and just try stuff or Replit, right? So I think horizontal, vertical, and then I think, you know, kind of open, closed. So there are tools that let you, you know, use it on their platform and you don't necessarily know what's happening. So that would be obviously like ChatGPT or Claude. You can't change the kind of rules under Pinninet. Also, you have to go to their website, you can't brand it, you don't really have much control over the privacy. And then more open tools are ones that you could put on your own site. You could add privacy into it, you could brand it. So that's what CHIP is, you know, we want to bring the power of AI tools like ChatGPT and Claude to your website,

• 18:02Speaker 02

  add privacy so the files stay locally, add your own branding, you can see the chat logs. So just a lot more control, obviously like Prediction Guard, same thing, right? Where you can bring AI into your own cloud. So a little bit more work, obviously with an open tool or you more power also you get more options. So I think that's kind of like. The lay of the land. And I think it's just like when you look at the internet broadly, like it started with text because that was easy to send across wires and then music because M three threes were smaller than video and then video. See the same thing with AI, right? Where it started with text and code because that's text heavy. Starting to get pretty good images now. Video is still coming, not quite there yet, but getting better every day.

• 18:42Speaker 02

  So I kind of see that evolution happening. Yeah. And I think what maybe is a surprise is that people thought the value was in the large language models. And I think what's become really clear the last month or two is it's actually going to be in the customer relationship and making the stuff easier to use. Deep seek is the model that came out of China a couple of weeks ago. And, you know, if I look at what chips AP like cost per API call is, it's gone down 90% in 18 months. Right? So just think about the, the value of these large language models becoming more commoditized. And then what, what is, what people are signing up for is like the experience of signing up and creating. So I can go to Replit and say, I want an app that, you know, is tracking my to dos and get it in a few minutes. It's

• 19:27Speaker 02

  all on top of the same power, right? It's all on top of chat GPT or Claude, just like chip, you can use any model underneath, but it's that end user experience, which maybe isn't so different than the web, right? There's protocols underneath, but you still use the browser that you like or the web app you like because of how it works. Not necessarily that it uses FTP versus something else, right?

• 19:49Chris Benson?

  Could you talk a little bit more about that end user experience, both the good and the bad? Because I, I think, you know, kind of going back to, to what we were talking about before, it, it's one of those barriers. And, you know, there's a set of people that are totally bought in across a whole bunch of different industries, but there's also a very large segment of the population that still really hasn't engaged. You know, they're hearing about it every day in the news and everything, but they're just intimidated and haven't done. So could you talk a little bit about the landscape of being on both sides of that barrier for different people?

• 20:23Speaker 02

  I mean, the biggest increase in use that we see with AI is putting it where people already are. So they don't have to learn a new interface, right? So if they can engage with AI via a Slack channel or via WhatsApp or via text message, like way easier, right? And so I think it's fascinating to see there's a lot of amazing UIs out there. But it's still like getting people there. It seems like time to value is really important with the tools. So like the faster you can show somebody an outcome. And that's, I think, where a lot of the new kind of text to app tools like lovable and Bolt are are really exciting for people because they can get something quick, which makes sense. I think that's kind of like how all UI is, is like, how do you get someone to the value quickest?

• 21:07Speaker 02

  I actually think like the default UI we are accustomed to with chat GPT is is not great. You know, like for someone to come in there and use. You know, it's interesting that chat GPT was a research project. It was not supposed to be a consumer app and it just became that on accident. And so I think there's a lot of improvements to the UI to come to make it easier for people to use. And you see those already coming into play where there's pre-built ideas, autofill, you know, connect to data sources. You know, the most common way people use chip is by duplicating an existing app. Right? So it's like solving that blank page problem is really important, I think, for any AI tool. So the easier you get people to motion is key.

• 21:48Daniel Whitenack?

  Yeah, I'm intrigued. You made me think of something so like for those that haven't seen chip and what Scott and team are building, you can go in and create individual assistants that, as Scott mentioned, you can kind of control and configure, make the way you want, connect the data sources you want. And often I think in my conversations in the past with Scott, I've heard him talk about how people are creating sort of proliferating these, right? You create one to do this and like one to do that and one to do this. And you clone this one to do that because it's not quite that, which is a different. It's a different paradigm than the sort of like, here's a chat interface. This chat interface is going to do everything that we that we want it to do. Could you talk about that?

• 22:33Daniel Whitenack?

  That element of it a little bit and what you've seen there? Because I I also see this on the business side, like when we engage customers, the kind of tendency, it seems from my perspective, is to say, hey, how are we going to build like our? Internal AI, right? And get it to do all the things that we want it to do. But it's like a single, in their mind, it's a single thing, right? It's like, this is our tool and it's going to be the tool to sort of rule them all. They're thinking very singularly in that way, which definitely does not seem to be kind of how people are engaging in the way they're building assistants in your tooling. Any thoughts there?

• 23:14Speaker 02

  I mean, I think the high level thought is the concept of software is getting turned on its head, where software is now an individual sport, not a team sport. You know, you think about, if you're the CTO even a few years ago, it's like, I have to do a lot of research, buy the right thing because everyone's going to use this. It has to fit the most use cases. We have to squeeze everything we can into one thing. And now it's flipped where every single person can build custom software within, you know, we say 60 seconds, right? So you would never build software to, I don't know, write a better introduction paragraph to a grant. But now like someone on chip will go build an app that just does introductory paragraphs for grant applications because it takes 60 seconds and it saves them three minutes every single time.

• 23:59Speaker 02

  And they do 10 a day and so it's 30 minutes. And, you know, we're seeing the average admin person saving 60 minutes a day on chip, going from 90 minutes to 30 minutes on admin work because they're building specific apps for their specific tools. So, you know, today I was looking at one that was getting IRS status from the IRS website, right? And putting it on to a spreadsheet. And it's like, nobody is going to go build a SaaS tool that just does that because the market is, you know, maybe 100 people or something. But with AI you can. And so there's definitely no need to have this like laborious, top-down purchase cycle when you can say, just try it. Like, does this solve one problem, two problems, five problems, 10 problems? Great. Like, imagine the power of every single person in your org

• 24:44Speaker 02

  being a web developer or a coder. Like, that's what it is now, right? And so now I don't have to bother our IT people or our developers. They can go do like the hard stuff, integrating with antiquated systems, right? Like getting our billing to talk to our web to talk to this. But for my job, I just have a file and I need to get something done. And like, I'm not going to bother our developer, but I'm going to be my own developer. And I don't know. That's a total flip, right? Or now we're not making decisions for the org. We're making decisions for Scott and I can just build it myself. So the only limiter again is is agency. Like just go, you have to go do it. Most people still won't, even though the tool's right there. But if they can at least try once, it's not as hard as they might think.

• 25:25Chris Benson?

  So it's a fascinating point you're making there with it. But it does change, even though you're talking about flipping the model over, you know, from kind of catering to the the business as a whole, to being able to cater to each individual contributor in the business. By doing that, I'm curious, you know, that that opens up a lot of possibilities for how you might run the business going forward. Do you have any thoughts on like what that does to the business? If assuming let's in a in a hypothetical world that you could get your entire workforce to engage in that way? What do you think that does for a business? And how how might if you were the CEO of a business, how might you operate in such a way to change that if you were just everyone's empowered with AI agents that they can make

• 26:09Chris Benson?

  in 60 seconds? What does that do for them?

• 26:12Speaker 02

  Yeah. And this is what Chip's trying to build. This is really my Arizona, like where we think work is going is we need an umbrella of of of safety so that our employees can do whatever they want without feeling like they're going to break something. Like right now, the fear of messing up is greater than the fear of missing out. And so we need to like get rid of that fear of messing up. So I always say, you know, like the FOMO is greater than the FOMO. Like, we got to get rid of the FOMO because people aren't taking action because they're scared. And so I think if I'm a company, what I'm doing is I have my five to 10 core apps. This is how we work. When you start at Scott, Inc., you're going to go through the onboarding chatbot. You're going to get the content creator that writes everything in our voice. You're going to, you know, get the data analysis

• 26:55Speaker 02

  that's going to analyze the spreadsheets in the same way. So these are the apps everybody uses. This is company standard. This is getting the laptop with pre-built software. And then underneath that now, you can duplicate or build your own to how you work, right? So you have this layer of company-wide apps. And then I have my Scott apps and maybe they're only visible to me. And a lot of times I might even cross personal and professional potentially, right? Where it's like, here's my workout schedule and my agenda builder for work and my, I don't know, grant writer tool. But since it's underneath this umbrella, we know that it's gonna adhere to privacy. Any personal information will be removed, so it doesn't violate any problems. And then the final piece is, yeah, we have the tools, but then we need that monthly or bi-weekly lunch and learn where like, hey Scott, what did you build this week?

• 27:40Speaker 02

  Oh, cool, let's just duplicate that one click and now send it to Dan. And Dan has similar work or, you know, new employee starts. They can look over my shoulder. It already, the bots already trained on all the history. It knows what to do, so they can jump in. And, you know, I always, I always say that AI really raises the floor. You know, like every new employee could start at average or slightly above average. You still need to raise the ceiling yourself, add that special spice, right? Your own ideas, but it's gonna make everyone on a whole quicker to get to work and higher, I guess, like higher average across the board. And I always tell, you know, the framework I always recommend is like the AI sandwich. Like just think about you, the AI interaction starts with you, the human, the bread on top.

• 28:23Speaker 02

  Then the AI is gonna do something, that's the meat in the middle. But then you still have to be the human on the bottom to take that output and to improve it, to share it, to repurpose it. And so I think a lot of new people get the bread and the meat, but they forget the bottom piece of bread. And so that'd be like the work I would do as a leader is, here's our tools, you can all use it and you're all gonna be good. Like you're not gonna have spelling mistakes, it won't be wordy, it'll make sense. But now how do you get better? And it's gonna be like adding your own spice on that last piece of bread. So that's what I would do for Scott Inc. So I think home run.

• 28:57Daniel Whitenack?

  And part of that too is like developing the muscle memory. So like for me, for example, the You know, we've been going through fundraising recently. There's always like the same set of questions that come up in diligence and in questions about the product and all this. And most of those have been answered like three million times now in some form. And you know, now looking back like, and you know, we've started to do this actually, but really what would be best is if we just had a little chat that had all of that preloaded into it and could chat over that.

• 29:38Daniel Whitenack?

  But at the time, it's like, oh, I'll just answer this email that's asking these 10 questions, right? I can bang that out really quick. But that, I guess there's a muscle memory thing there. And then there's a, there is some barrier to overcome to configure the system for future benefit, right? That you might not see there. So I don't know. Yeah. Any, any suggestions even in your own personal life where you've kind of come over?

• 30:06Speaker 02

  I mean, we did, we did the same thing, right? Like we did a raise with Chip and we built a Chip chat and it was trained on all of our, you know, slides and everything. And people still want to talk to you like, it doesn't mean that they don't get a human.

• 30:19Daniel Whitenack?

  Right, exactly.

• 30:20Speaker 02

  But it gives them the option and like, you know, the data we're seeing for our users using Chip for, for like customer support, like a chat bubble sort of use case, 70% of them are not clicking the talk to a human button. Like they just want to know, what are your opening hours? How much does it cost? Who are you? Like, just give me the facts. And as like a busy parent, I get that, right? Like, I don't want to make phone calls because I know it'll take five to 10 minutes versus a minute if I'm doing it myself. So, so I think there's that aspect of like time efficiency. And it is changing habits of like going somewhere else or like you said, taking core info and putting it into a repository. What we found most helpful is we have something called dynamic knowledge sources. So if it's a spreadsheet or a folder on Google Drive or OneDrive, anything that gets

• 31:03Speaker 02

  added into those places is automatically added into your agent. And so I think with businesses, it's important to think about that flow of information and minimizing as much like documentation work as you can. So we always put everything into Notion or Confluence or Google Sheets or Google Docs. Make that your. Hub that is fed into the AI. So everything that you put in that place gets automatically added into your FAQ bot or your, you know, marketing assistant bot or whatever. So I think that's, that's key is like, you can ask people to do it, but even better is like not to require more work or even changing behavior because we know that's the hardest part.

• 31:44Speaker 02

  So maybe it's a BCC email that goes into a spreadsheet that's automated, right? Or, you know, something like that. So you can kind of decide. The way we do it is we actually look at our chat logs of people engaging with chip and find the answers that are going unanswered or don't have a great answer. And then we add those things in once a week into our chat so that it improves for the things people are asking for, rather than trying to solve for hypothetical edge cases. Yeah.

• 32:09Daniel Whitenack?

  Well, Scott, we've, we've kind of, we've, we've talked a little bit about chip. I've described it a little, a little bit. I'm, I'm wondering maybe for, you know, you've been on this journey of kind of trying to build this easy to use AI tool. Along that journey, have you found, I'm sure you tried various things that did work and didn't work and certain things have been difficult and certain things have been easier. As you reflect on that kind of, as a founder of, of an AI company, trying to build an AI tool, any things that you'd want to highlight in terms of things that were kind of key insights or, or bumps along the road that, that in retrospect you look at and kind of

• 32:54Daniel Whitenack?

  make sense or, or anything like that? Cause I think there are a lot in, in our audience that have, have maybe ideas for, for things out there. Yeah.

• 33:03Speaker 02

  No, it's amazing. There's so many, I think, I'll take like a non-obvious one, which is we, we were pretty early on focused on building community. So we have over 20 chip chapters around the world, people teaching one another AI, fairly active Discord. That's been invaluable because those are the people who are bringing back problems and ideas and being able to build towards actual customer questions is, is so important. And a lot of times customers don't have time or interest in. You know, giving you feedback, what you need. And so what we've done is like every two weeks or so, basically having free workshops to try to educate our users and anybody. And that's really built a relationship, I think, where we know these people by name.

• 33:46Speaker 02

  We know where they live, what they do, and it makes it a lot easier for them to be like, yo, can you build this thing? I need it for a pitch on Friday. And we're like, yeah, for you, of course, because you're contributing, you know? So it's building that, uh, building relationships and it doesn't have to be hundreds, right? This can be dozens of people who love you. And that's how you really start is like a strong foundation. So I think that one's non-obvious. Um, I think technically something that we found maybe an accident and we're trying to lean into now is riding the wave of other people's innovation, you know? Like you can only build so many unique pieces and you need to be on top of other parts of the tech stack. And so, you know, chip is built on top of large language models. So as Anthropic and OpenAI build better models, chip gets better.

• 34:30Speaker 02

  And for a lot of our users, they think chip is doing that because, you know, we are their front door to AI. And so as the models get better, chip gets better and their experience gets better. We partner with folks like Prediction Guard who help us provide better privacy and security. Right? And so we could go spend six months trying to build that, but now we've lost the whole point of what we're doing, right? And so what is your forte is really important. One thing that has really recently that we kind of focused on is Anthropic has a new protocol called, uh, what is it? Model Context Protocol. Um, it's basically an easy way to connect APIs into AI tools. And so that's another example of like, we've been building one-off APIs to all these different tools. And now it's like, wow, there's this whole world that's built towards this standard.

• 35:14Speaker 02

  And if we just tap into that, now we can, again, get better, the more the open source community contributes. Um, so I think that's really interesting to look at where are the areas that will move quickly, that you can ride that wave, and then where do you want to be a differentiator? And you can kind of draw your line wherever the right place is. Um, but probably don't try to draw it on all of them. Like pick the ones you're best at. Yeah, I think those are, those are a few, and I think just the power of small teams now. I mean, you read that a lot of places, but. You know, our CTO Hunter, who is just like a beast with AI coding. And it's like, I know our output compared to some legacy teams is just vastly greater. And so I wouldn't underestimate if you're a solo founder, you got a team, you know, we have a couple of chip users.

• 35:58Speaker 02

  There's a guy named Chuck in Colorado who he's building a million dollar, one person agency. And he's almost there, right? And it's all built with AI automations and he's conducting everything. There's a lot of potential out there. So I would encourage you when listening, like finding a co-founder or a team is really, really hard, but you don't have to wait. Like you can do a lot on your own.

• 36:17Chris Benson?

  I'm curious, you actually started to get in for a second to the next question I was going to answer. And that was, you mentioned like privacy and security and partnering with Prediction Guard for that. As you're thinking about these, these different concerns that weigh in on various industries. And, you know, there'll be, you know, legal concerns, things like, you know, HIPAA in the medical world. And every industry has its own set of concerns that are kind of external, but are binding the work in those areas. And as you are kind of, kind of unleashing people's potential with the work that you're doing.

• 36:56Chris Benson?

  Those kind of have to find some sort of balance. How are you thinking about the constraints versus the unleashing? That we talked about and finding a balance so that people are unleashed while they're still having to be held to account. You know, by whatever those constraints in their industry is.

• 37:12Speaker 02

  Right. Yeah. I mean, I think regulation is always going to trail the innovation. And so I would say as a company, as an individual, like look at yourself first before worrying about the regulatory environment. You know, I think about privacy pyramid is what we tell our customers, like the bottom of the pyramid. The first thing you should do is just think about what are you okay sharing and not sharing and just tell people. Again, like FOMO is greater than FOMO. People will not take action if they think they're going to get in trouble. Even if it's hypothetical, like not real, like, I don't know, that fear from elementary school like sticks with us, you know? And, and so the first thing you have to do is remove the fear. And the best way to do that is just to say what the rules are. As long as people know the rules, they'll work within them.

• 37:58Speaker 02

  But if they don't know what they are, they're afraid that whatever they do will get them in trouble. Right? So, hey. Just don't upload customer data. Like, that's our rule. Great. That's a great place to start. Now go do anything else. Or, you know, no customer data and don't integrate with these files. Great. And the second level of the pyramid after, you know, kind of just best practices internally, is then going to be like, human protection error, I call it, which, you know, one thing Prediction Guard offers as well, which is like encrypting pieces of information that get added that shouldn't be, right? So if I add a phone number or, you know, a social security number or something like it gets removed for me because I made a mistake. That's fine. Like, we make mistakes, but best practices, and then cover other people's mistakes up as they make them. And then I think the top of the

• 38:43Speaker 02

  pyramid is where you actually say, you know what? Let's put it in our own environment. So that way, if we can share whatever we want without having to worry. And, you know, that's where you can run an open source, large language model in your own cloud infrastructure, whatever you share is in your cloud infrastructure. So, you know, some businesses have to do that. So if you are in finance, healthcare, like, you're probably going to want to do that anyway, just for regulatory reasons. Some people want to do that because they know they're going to be sharing data that might be sensitive. But I think for most of us, like, to get started, just follow that basic best practice of like, think about it before you share it. And if you're working with a team that might make mistakes, or contractors who aren't following your rules, like, add in that second level of, like, human error protection.

• 39:24Daniel Whitenack?

  Scott, as we, as we kind of get nearer to the end here, I'm wondering if you can maybe share just a few standout use cases of maybe things that you've seen people do with Chip that have either surprised you or stood out in a, in a way like, oh, I didn't expect people would do this, that, you know, or, or things that are like, oh, I didn't even know, you know, I built the platform, but I didn't even know that was possible.

• 39:53Speaker 02

  Every day. That's my favorite part of Chip and AI generally, is like, we really are building the tools and we don't know how people will use them. And it's so crazy to see what people do with it. And, I mean, the most common use case is I always say like, there's kind of five areas that people use all the time. It's like operations, marketing, sales. I call it company search, like finding stuff in your Google Drive, basically. And what's the last one? Like data analysis, like reviewing financials and things like that. So those are like the most common. But in terms of like fun, weird ones, like we had somebody who launched a Canadian tariff checker. And so like, as the tariffs on Canada were released, you could actually search any product and it would source like where they were coming from and tell you what the change

• 40:36Speaker 02

  in price would be. That was like totally interesting. One of my favorite use cases, a guy named Tyler Hanson, he's in Sioux Falls, South Dakota, and he runs an HVAC company. And he put in all of the training manuals for all of the equipment that they service. So then his technicians are on the, on the ground. And instead of having to like be in the bathroom, watching a YouTube video, which I know has happened when my HVAC guy comes, right? He's like actually learning how to do the thing that I asked him to do. Like, they can actually pull up the specific model via their chip chat and get instructions on what to do and how to service it and parts. And that one's really fun. There's a contractor out in Washington, he uses it to create supply lists. So he just puts in square footage and what people are going to build. And then it'll spit out like how much wood he needs, how

• 41:19Speaker 02

  many nails, like whatever else. Again, like things I know nothing about. A lot of people doing it for like finding HR policies, finding, let's see, there's a car dealer that's using it to find cars to purchase, like to then resell, right? So like searches through AutoTrader and, you know, Craigslist or wherever else to find vehicles. Just so many things, right? And every day I'm encountering new ones that are so fascinating. The, the fun part is we integrate with, you know, APIs and webhooks. So really like any tool can get pulled in. And a lot of times chip ends up being a front end to an AI tool that's talking to their software. So chip becomes the way they communicate, but then it's pulling their own data. So that's

• 42:02Speaker 02

  super fun. Personally, I have a Scottbot, you know, that's the one I use every single day. And so like I can write things very quickly and remember people that I've talked to. So it can like brings in past conversations and. So that helps me quite a bit. So yeah, those are a few random ideas. I haven't built the West Lafayette tour guide yet, but we do have some travel, travel AI tools out there. So I bet we could do that, too.

• 42:26Chris Benson?

  So very cool. And while you're, while you're building that tour guide, I might give you a location or two as well.

• 42:31Speaker 02

  Okay, there you go. Yeah, that's awesome.

• 42:33Chris Benson?

  So really cool use cases there as you like, that's got to get you thinking about like the possibilities. So you know, you come at it with your own mindset and the things that you have. Your, your customers are teaching you every day about what the new possibilities and boundaries might be. So where does that, where does that take you? Like when you are, are, you know, you're kind of done for the workday, your brain's decompressing and, but you're still kind of, you know, just working on things. What, what are, what's going through your head about like, where could things go with this? You know, you, you take what you're driving and, and the, and the folks you're working with are driving, you're taking what your customers are showing you that you never

• 43:15Chris Benson?

  thought about. And that's got to leave you with some pretty cool ideas about what the future might hold. But can you share some of those ideas with us?

• 43:22Speaker 02

  Yeah, I think, I mean, I, I reflect at the end of the day in a lot of ways because I have four kids that are 11, 9, 7, 3. And I just really try to think about like, what does society look like when this is more present? And, you know, what does education look like? I spent a lot of my life in education. We work with a lot of schools who use it for tutors and advisors and, you know, what's the value of a credential saying you know something when the pace of change is like way faster than four years, right? I think ultimately, you know, I imagine this technology has to fade away from being AI and just being a part of what we use. And it helps us lean into the things that make us weird. You know,

• 44:05Speaker 02

  I think about AI is the world's best cover band, and it needs like the originals to cover. And so I think it's really forces us to be more unique as individuals and, and create something new rather, you know, we're going to use AI for a lot of the, the quick answers and it's going to be average. It's going to be the middle of that bell curve and, and that'll be fine for most work. But again, we have to raise the ceiling ourselves. And so I think it makes me feel like I want my kids and hopefully myself to like, Just get good, really good at whatever weird, interesting thing we care about. Yeah, and I, man, I don't know. I think agency, again, like, I keep coming back to that. But how do you instill a lack, like a fearlessness in people? Because it feels like, first of all, most people aren't aware of the pace of change.

• 44:50Speaker 02

  And as they become aware of it, it's either I'm scared, I'm gonna back away, or I'm gonna lean into it. And I think we just really need to lean into it. And I don't know, I think it's exciting because I'm in Fargo and I couldn't, you know, learn to be a nuclear physicist in Fargo, right? But now I could, like, I can easily go down that path and learn what I need to connect with the resources, you know, showcase my work. And this has kind of been my dream since my first company in 2009 of like, really giving, giving anyone, wherever they are, a chance to, to build. And AI is just like the next step in that process. And I know a lot of people still will find reasons not to, but it's gonna be just on that agency piece, like, you can.

• 45:34Speaker 02

  And so I don't know, I think a society where everybody has a chance to, to build and create is incredibly exciting. It's gonna be more competitive, you know, everyone around the world has equal access to the same models as NASA and, you know, like the Defense Department. Like, it's kind of wild that you can log into these things for free and have the same power as everyone else. So that's an opportunity if you, if you take it. I think, I saw, there was a recent study, um, the World Bank did in Nigeria, and students who are using ChatGPT as a tutor for six weeks had the equivalent of two years of education. And it's just, so many of our problems are problems of access. And I think a lot of those access problems go away. And then what happens when another, you know, one billion people come online with education who don't have it now? Like, that's just better for us all.

• 46:19Speaker 02

  We can come up with really exciting solutions to our problems.

• 46:22Daniel Whitenack?

  Well said. Yeah, that's a, that's a great way to end. Thanks for, thanks for joining, Scott. I encourage everyone to go, uh, create your first Chipp chat on, uh, chipp, C-H-I-P-P dot AI, um, and, and have some fun. Uh, explore those, that weirdness, as, as Scott put it, I, I love that. Thanks for joining, Scott. It's been great to chat.

• 46:43Speaker 02

  Great to be here, guys.

• 46:51Speaker 00

  All right, that is our show for this week. If you haven't checked out our Change Log newsletter, head to changelog.com/news. There you'll find 29 reasons, yes, 29 reasons, why you should subscribe. I'll tell you reason number 17. You might actually start looking forward to Mondays.

• 47:12Speaker 05

  Sounds like somebody's got a case of the Mondays.

• 47:15Speaker 00

  28 more reasons are waiting for you at changelog.com/news. Thanks again to our partners at fly.io, to Brake Master Cylinder for the beats, and to you for listening. That is all for now, but we'll talk to you again next time.