The $3 Trillion AI Coding Opportunity

Originally published on the a16z Infra podcast. We're resurfacing it here for our main feed audience. AI coding is already actively changing how software gets built. a16z Infra Partners Yoko Li and Guido Appenzeller break down how "agents with environments" are changing the dev loop; why repos and PRs may need new abstractions; and where ROI is showing up first. We also cover token economics for engineering teams, the emerging agent toolbox, and founder opportunities when you treat agents as users, not just tools.

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

• 0:00Guido Appenzeller?

  AI coding is the first really large market for AI.

• 0:05Speaker 01

  When do we say this is our agents? We just at the end of the value chain, we're like, does this work or not work? Click yes or no. Agents, more than ever, need an environment to run these things. Context engineering for both humans and agents.

• 0:20Guido Appenzeller?

  Every single part of it is getting disrupted. It's not that there's just somebody writing code, like your classical developers being disrupted, but everybody along the value chain.

• 0:30Erik Torenberg?

  We're resurfacing an episode from the AI plus A16Z podcast that we think deserves a wider audience. A16Z partners Yoko Lee and Guido Appenzeller make the case that AI coding is the first truly massive market for AI, potentially worth trillions. They break down what's changing in the dev loop, where ROI is showing up first and what founders should build next. Hope you enjoy.

• 0:54Guido Appenzeller?

  So Yoko, we just launched this, I think, amazing new DevStack for the AI coding environment. And I'm really excited about this.

• 1:02Speaker 01

  Yeah.

• 1:03Guido Appenzeller?

  I mean, and let me start with a very high order pitch, why I think this is so incredibly exciting. I think AI coding is the first really large market for AI, right? I mean, we've seen there's a ton of investment has flown and the question now, to some degrees, is about the value, right? Why are we doing all this? AI coding can create an incredible amount of value. If you think about this, we have about 30 million developers worldwide, roughly, right? Let's say each of them generates $100,000 in value. In the United States, it may be low because many of them get paid a lot more, but internationally, it might be a little high, but I think in order of magnitude it holds. So in aggregate, the value we're creating here is about 30 million times 100,000, so $3 trillion.

• 1:44Speaker 01

  I will argue even more because that's just developers, but then there's also people who are development curious. They were not developers, maybe they're, I mean, design engineering now is a big thing. Every designer should just, you know, write code.

• 1:57Guido Appenzeller?

  Product managers.

• 1:58Speaker 01

  Exactly.

• 1:58Guido Appenzeller?

  Yeah, I mean, there's only this effect. But if you just take the $3 trillion figure, that's about the GDP of France. So the claim we're making here, as crazy as it sounds, is that we're saying the entire population of the seventh or eighth, I think, largest economy on the planet generates about as much value as a couple of startups that are reshaping the AI software development ecosystem plus the LLMs underneath.

• 2:20Speaker 01

  Yeah.

• 2:21Guido Appenzeller?

  That's crazy.

• 2:21Speaker 01

  And everything we see and touch and use nowadays are all software.

• 2:26Guido Appenzeller?

  That's right.

• 2:26Speaker 01

  Yeah.

• 2:27Guido Appenzeller?

  So we've, software has disrupted everything in the world. And now, software itself is getting massively disrupted.

• 2:35Speaker 01

  Totally. And then, what you mentioned in the blog post is really interesting, just because we now are more capable at using L M to generate code and coding and produce software. But then, as a result, it's not like there's less jobs. There's actually more and more software is being produced. Before, maybe it's a SaaS service catering to hundreds of people's needs, two thousand of people's needs. Now you can really just vibe code things, software by one, for one.

• 3:00Guido Appenzeller?

  Yeah.

• 3:01Speaker 01

  Yeah, I vibe code.

• 3:01Guido Appenzeller?

  I do that, yeah.

• 3:02Speaker 01

  Exactly, you do that, exactly. And then I vibe code my own email filter. You know, I don't do that so much of using L M to reply to my email. But I have a filter where I categorize the labels, things like that.

• 3:13Guido Appenzeller?

  Only to some emails.

• 3:14Speaker 01

  Only to some emails.

• 3:17Speaker 01

  So the first question becomes like, how do you think the development loop is shifting?

• 3:22Guido Appenzeller?

  I think the answer is complex. And very frankly, it's so early, I think, in this AI revolution, we don't have the full answer yet.

• 3:29Speaker 01

  Right.

• 3:30Guido Appenzeller?

  Right, but, I mean, we have our little stack, our little software development lifecycle post AI in the blog post. And I think the biggest learning from that is probably that every single part of it is getting disrupted. It's not that there's just somebody writing code, like your classical developers being disrupted, but everybody along the value chain.

• 3:46Speaker 01

  Yeah.

• 3:46Guido Appenzeller?

  Is getting disrupted.

• 3:47Speaker 01

  What's the most surprising part for you? What's the most disrupted field today in coding? And what do you think AI will come after next?

• 3:55Guido Appenzeller?

  So I think, well, we've seen the biggest growth, I can say to say, in the classic sort of coding, IDE-integrated coding assistants or more gentle coding assistants, right? The cursors and devins and GitHub Copilots and then Claude Codes of the world, right? I think that's where we see the most traction, where we see incredible revenue growth. I mean, I want to say that segment possibly has the fastest revenue growth of any startup sector we've seen in the history of startups. Which is again incredible statement. So I think this is currently the vanguard, right? And everybody's aware of it. We're seeing billion dollar acqui-hires or takeover offers. So that's incredibly vibrant sector. Now, which one is next? That's a really good question.

• 4:34Speaker 01

  So to be very specific, in a blog, we wrote about the basic loop. The basics is you plan, you code, you review.

• 4:42Guido Appenzeller?

  Yeah.

• 4:42Speaker 01

  Where does LM coming in? Where do you think more of the loop would be disrupted? Do you think the loop will still be the same as what we used to have, the basic? Or do you think it will look very different?

• 4:53Guido Appenzeller?

  I think at this point, it's very hard to speculate about the end state. But if you, let's assume you've seen the first email.

• 4:58Speaker 01

  By C tape, yeah.

• 4:59Guido Appenzeller?

  No, I'll get there. But if you looked at the first email sent over the internet, right? You can sort of predict that probably we'll have websites and these things. Maybe if you're good, you can, right? But you know, as saying like, hey, the net effect of this is that everybody can rent out their house and compete with hotels. And this is going to be the biggest hotel company on the planet. You'd be like, well, that's a little farfetched. But now we have Airbnb, right? So also the secondary effects, I think, are really hard to guess. My current hypothesis is I think we'll still have software developers. I think they're not going anywhere.

• 5:26Speaker 01

  Yeah.

• 5:26Guido Appenzeller?

  Right? I think what they do will look completely different.

• 5:30Speaker 01

  Right.

• 5:31Guido Appenzeller?

  Right?

• 5:31Speaker 01

  Yeah.

• 5:32Guido Appenzeller?

  I think the CS education, frankly, any CS class taught today at any major university is probably best seen as this historical relic from a bygone time, right? I mean, if you look at the best of breed startups, what they're doing, the loop that the developer is in looks so different from what you did before, right? You have multiple agents that you're prompting, that you're telling things, you know, you pull that back into UI, you're trying to understand what they did. You're trying to put them back on the rails. So a lot more thinking at a higher level. I mean, all of coding sort of has been higher levels of distractions, but I think we're making a huge leap here. So how it's going to look like, I have no idea. My gut feeling is we'll probably have more developers. Look, this basic. Plan, execute, cycle. There's probably going to be some flavor of that's still around, is my guess. But what do you think?

• 6:16Speaker 01

  So one of my top questions is that would this be a step by step loop? Or would this meshed into just one step? So one example is if my agent is writing a code, do I still need to review it? Or do I have another agent that just reviews it? If it's the same agent, like implementation detail is one thing. You can separate out the agent generating code from the agent reviewing code. But then if it's all agent, both generating and reviewing code, is it just the same step? Do we actually disaggregate the step by step process and have human in the loop? Whereas if as a human, I write code and I want agent to review code, that makes sense to me. So I do wonder, when do we pull out something as this is an individual tool

• 7:01Speaker 01

  and individual step an agent takes care of? And when do we say this is all agents? We just at the end of the value chain, we're like, does this work or not work? A click yes or no?

• 7:12Guido Appenzeller?

  I think the time periods over which an agent can work autonomously will get longer. You know, still, if somebody says, look, I want to write a complete ERP system for my multinational enterprise, go. There's no way I could imagine that it'll just run and back come software that actually fits the requirements. And in part, I think it's a problem that models are still very, very far away from being able to run autonomously for that long. But the other problem is, let's assume this was an all human team. We wouldn't understand all the challenges yet at the beginning, right? We'd have to revisit the design. We have to revisit the architecture that has cost implications and so on. So at some point, you sort of need to go back to the architects and the product managers and say, hey, we had a plan A, didn't quite work, or we have found new challenges. So here's our updated plan A. Is that the, you know, or a plan B, right?

• 7:57Guido Appenzeller?

  Is this what you want to do? So I think the loop will still be there. The time scales will probably change, but yeah, it's very, very hard to guess right now.

• 8:03Speaker 01

  Yeah. Another thing we start to see more often is contrary to how much humans need to come and intervene the loop. We actually start to give agents tools for them to know what they have to do. One example is this loop I see so often, like the agent wants to implement, say, Clerk off in their app. Now they need to go to MITLify and contact seven to say, what is the latest version of Clerk? And how can I implement it correctly and in what file? I'm not going to copy paste it to cursor or give it to the agent because I'm too lazy as a human now. The agent should be able to call the API themselves to put stuff in the context to make it work. This is just one example of what behavior change we're seeing.

• 8:48Speaker 01

  Because before, as developers, we're so used to go back to the docs and refer to the docs and tell the agent what to do. Now agents can obviously just-

• 8:55Guido Appenzeller?

  We cut out the middleman.

• 8:56Speaker 01

  Right, we cut off the middleman. I don't need to route all these requests for the agents anymore. I think there's other examples, which is verification. As a human, before when I write code or review other people's code, I pull out the code. And then the first thing I do is actually not to review because I don't like reading code. I'm not a human compiler. The first thing I do is to fork the change and see if it still works. If it doesn't work, I just do not review it. Nowadays, there are opportunities to give just agents a native environment to first see, does this work? Does the UI still look good? Do all the requests still check out?

• 9:32Guido Appenzeller?

  Yeah, absolutely, yeah.

• 9:33Speaker 01

  Did it break my build before the human needs to come in and review? Maybe that manifests itself in part of the local development process. Maybe it manifests itself in the PR review process. But in any case, now agents, more than ever, need an environment to run these things.

• 9:49Guido Appenzeller?

  When I used to write something just for myself, like a little script I need somewhere in the past, I usually didn't include unit tests, right? For production code, it's different, but for just personal things, you know, it's like, yeah, that's a single developer, I know what I'm doing here. With agents, I now start including unit tests because they're so much easier to write. And they allow an agent, as you said, right, to understand if the changes that they did, did broke anything else, right? And they may not have the context, how this originally was built anymore, and then easily to adjust to perform. So that's a good value.

• 10:15Speaker 01

  On a grand scheme of like, how much economic value this generates, where in the value chain do you think it's growing the fastest? Like, you see-

• 10:24Guido Appenzeller?

  Right now?

• 10:24Speaker 01

  Where the agents is producing so much more value than other areas. But what are the areas you feel like would be the next takeoff?

• 10:30Guido Appenzeller?

  So, look, I'm, I've, I'm talking about a hundred or so enterprises about this per year. Just when we, you know, take our portfolio companies to them as potential customers. What I'm hearing from them is that the number one use case in terms of ROI right now is legacy code porting, right? It's not super surprising. Like one of the first papers in the space from Google, right? They, they wrote a fantastic paper on, uh, you know, where they detailed on, you know, just doing very mundane things like replacing a lot, a Java library across a very large code base, right? Not like millions of lines of code, right? So it's a very logical base.

• 11:01Speaker 01

  What do you consider as legacy stack? What do you consider as new stack?

• 11:06Guido Appenzeller?

  It totally depends on you, but I mean, for the banks, it's often COBOL or Fortran to do Java.

• 11:10Speaker 01

  Oh, COBOL. I haven't heard that word for a long time.

• 11:14Guido Appenzeller?

  You know, I, I actually wrote COBOL code once in the nineties. It probably dates me at this point.

• 11:19Speaker 01

  Okay, actually, how are L L M's with COBOL?

• 11:22Guido Appenzeller?

  They're apparently extremely good. Um, so

• 11:24Speaker 01

  That's a surprise.

• 11:25Guido Appenzeller?

  So here's the thing, right? One of the hardest thing, if you implement code with L L M's, is just getting the specification precise, right? If, if I can specify something very precisely, then usually the L can do a good job at implementing it. So many of these companies do is they take legacy code, they have an L L M, right? A specification that fits the legacy code. And then they say, re-implement the specification and you may look at the code if there, you know, as a tiebreaker, if something is not clear, right? And that seems to work incredibly well. So I'm, I'm hearing that today, um, I heard from several sources now that you can get, uh, you know, about a two X speed up versus traditional, um, processes for that, right? And, and that's amazing, right? And what this has led to is that actually of those enterprises that I've talked to, the majority says they're currently accelerating, like at least the, the

• 12:10Guido Appenzeller?

  majority of those that are sophisticated about this, they're saying they're accelerating the developer hiring, right? We don't know if this is a long-term trend, but right now they're basically saying, look, because we found so many low-hanging fruit type projects where with a, with a little bit upfront investment, we can then save infrastructure costs. And that's super exciting. So I'm, I'm, you know, what this will mean for the, you know, how, how much long is the mainframe business going to be around? Or, you know, I don't know, but, but it's, there's definitely a shift there where suddenly legacy code migration is much, much easier than it was before. Uh, you know, is, I think that they'll change a lot of the dynamics in the sort of classic enterprise software space.

• 12:42Speaker 01

  That's interesting. I do wonder if we will get new mainframe code because now, because before no one knows how to program those things.

• 12:50Guido Appenzeller?

  That could also be, yes.

• 12:51Speaker 01

  Right. And now you realize you can program mainframe using natural language.

• 12:56Guido Appenzeller?

  Yeah, totally.

• 12:56Speaker 01

  So another possibility is that we get renaissance of like the underlying legacy, uh, coding languages.

• 13:03Guido Appenzeller?

  It's crazy to me how versatile these coding assistants are, right? I mean, we're seeing them write CUDA kernels.

• 13:10Speaker 01

  Yeah.

• 13:10Guido Appenzeller?

  Which is like, that is difficult stuff to write by any metric.

• 13:13Speaker 01

  Yeah.

• 13:14Guido Appenzeller?

  I've tried them with a language which basically has no usable training data set. And they're still able to sort of abstract, you know, with a couple of examples, you know, of how the code will have to look like. It's not perfect, but so I think it's a very, very broad technology for sure.

• 13:30Speaker 01

  Recently, just like what we were talking about before, code reviews. Because like to your point, LLMs are so good at coding and generating code. Sometimes it's beyond our comprehension. We'll take more time to review the code than the coding agents. Like, that's not a controversial opinion, just like how it is.

• 13:50Guido Appenzeller?

  That's the reality.

• 13:50Speaker 01

  Yeah, so it does make me wonder how our development chain and steps are gonna evolve from here. How are we going to do PRs when we can't possibly review thousands of lines of code as humans? So does that mean the right abstraction now is still code? Or does it mean the right abstraction now is for us to review plans? If that's the case, is GitHub review still the right abstraction for that?

• 14:17Guido Appenzeller?

  I think there's still a role for review in general. I think the question is, will humans do the review? Like, right now, most of the code then LLM generates. You know, unless you're, if you're deep in vibe coding territory, you're just like, oh, this is a one-off. I'm, you know, just want to try something out. Maybe then you don't review it, you just hit accept and hope for the best. But anything that, you know, anything else, you do review the, I still review the code line by line. You know, that said, we're starting to see really good tools that plug into your backend system, your GitHub.

• 14:48Guido Appenzeller?

  Whenever a pull request comes in, they analyze it, they comment on it, they point out security vulnerabilities, they point out that the spec is different from the implementation. You know, they, they point out that this creates dependencies which may not be desired. They, you know, enforce coding guidelines, which is very, very powerful, right? I haven't met anybody yet. Tell me if you have, but I haven't met anybody yet who basically has said we're going to rely purely on AI to review code, right? We're just, anything can go in if the AI checks off on it. But I have seen, for example, companies that are saying before we had two developers review code and now it's one developer, right? And or, you know, cases where basically just the AI hangs out in the sort of in the GitHub, you know, discussion and comments on things, you know, if there's, if they, you know,

• 15:32Guido Appenzeller?

  so you can basically delegate tasks, like, can you look at this? It was like, are we using this library somewhere else? So basically, you can certainly have somebody who can, who can help with these tasks.

• 15:40Speaker 01

  My hot take on this is actually if PR is supposed to give us a context as developers on what did these other coding agents or my colleagues, you know, change that I should be aware of, I don't think code, reviewing code is the right abstraction.

• 15:57Speaker 01

  Because it might be feature level, it might be performance level, so now, now it might just become a white liner. This, you know, this agent improved your CUDA implementation. And I, maybe I don't even know CUDA, but I know the improvement when I can verify it. So the question for me is, do I still need to go to the PR review every line? Or should I just be given, you know, two sentences to know how this works and an environment to test it out? And that would just merge it.

• 16:28Guido Appenzeller?

  If it's the right two sentences that works?

• 16:30Speaker 01

  Yeah.

• 16:30Guido Appenzeller?

  Will the, will the L M always pick the right two ones? I don't know yet.

• 16:34Speaker 01

  Yeah, that's true. I mean, if you give it an environment, the question is like, can you verify the environment against the two sentences?

• 16:41Guido Appenzeller?

  Yeah.

• 16:41Speaker 01

  If the answer is yes, that would be easier to solve. If the answer is no, you know, that's harder.

• 16:46Guido Appenzeller?

  But I think there's a bigger picture here, though, which is that L L Ms are also very good in generating the documentation and description for the code, right? So when I use something like Cursor for Coding.

• 17:01Guido Appenzeller?

  And you know, it generates code. I often ask it afterwards and now take the internal documentation updated. Because the internal documentation is important for me, but but also for for the coding agents. Because they want to be able to refer to it. You don't want every single time to take the entire code base and stick it into the context window. It's massively inefficient and slow. So if instead you can just say like, read this document that'll explain to you the class hierarchy. Right. And then based on that, implement this new subclass or so it's much, much faster to do. I think I think there's a real opportunity there to get to much better documented code than previously.

• 17:32Speaker 01

  Right.

• 17:33Guido Appenzeller?

  You can almost. A compiler is great and then it takes a high level abstraction, translates it into a lower level one. But now we have the ability that if somebody had in hand optimized the low level one, we can use that to update the higher level one.

• 17:46Speaker 01

  That's true. What is the new compiler in the age of AI?

• 17:50Guido Appenzeller?

  L M's are sort of a compiler, right? In some way, they take a high level, higher level description and and filter down. I mean, I think the.

• 17:58Speaker 01

  I guess what's missing is it doesn't have a natively have an environment compiler give you, does this work? Or does it does it compile or does it not compile? It doesn't tell you, does this work, which is a very subjective thing.

• 18:09Guido Appenzeller?

  I think it's right, right? And like a compiler enforces is a strict enforcement of certain things, right? You can, I can rely on that. I don't know. I'm using Rust and things will be typed, right? So that's that's a huge step forward, right? And I can exclude certain bugs with an L M. That's not the case, right? The now that said, you sort of wonder, is this just an initial thing? Is this just a, I mean, L L M's over time, like for example, we can give L L M's tools that allow them allow them to syntactically parse code, right? And then now suddenly they can start reasoning above code. They can ask, are we sure that the representation of object X is the same across these different, you know, modules? Even if it gets serialized in between or something like.

• 18:54Speaker 01

  I don't mean to make this episode disaggregating GitHub, but GitHub is so central to so much of our workflow, right? Everything goes through it. From the social aspect of discovering what other developers are doing, to distributing the software, you can download things, to keeping track of what you have changed, Git.

• 19:12Guido Appenzeller?

  Whole integration with the build system, the backend, yeah.

• 19:15Speaker 01

  Yeah, so now, another interesting thing we start to see is people use Git repos very differently now. Before, it was like, oh, humans will make some changes, we'll commit it, and then other people will see it. And then we, you know, open the PR, people see different revisions. Now, agents are doing so many changes. It's kind of counterproductive to commit everything. And then repos usually have very low rate limits because it's designed for humans to use. So now the question becomes, what is the new repo abstraction? Like, that handles distributed infrastructure, but also caters to high-frequency commits.

• 19:59Speaker 01

  And sometimes, when agents doing these commits, they don't even want to preserve this forever. It's more like an intermediate step so that they can explore five different paths, you know, but then revert to any of them. And then when they're happy with it, they commit back to GitHub. So one thing I've been playing around with is there's this company called Relays, and then their docs has this repos feature where you can give a repo to the agent, the agent will, you know, commit stuff like very high-frequency, kind of works really well with by-coding agents. And that has been such a great experience. And in the wild, I start to see people building this internally, too.

• 20:39Guido Appenzeller?

  It makes total sense. Look, if we completely change how humans write code, or we're shifting from humans to delegating the most of the writing to agents, I think it would be foolish to assume that the underlying services that were a good fit for the human world are still a good fit for this new agent.

• 20:56Speaker 01

  Yeah, for sure.

• 20:57Guido Appenzeller?

  Right? That's almost certainly not the case. Well, we can do better in our case. I think you're completely right, right? For source code repositories, we want something that's much more real-time. Honestly, I mean, imagine, let's take this to the limit, right? Let's assume I have now, I'm just me personally, running hundred agents in parallel that all try to implement features on my codebase.

• 21:15Speaker 01

  Yeah.

• 21:16Guido Appenzeller?

  We probably need some kind of coordination mechanism between them. Then all, they're all not trying to edit the same file.

• 21:21Speaker 01

  Oh, for sure.

• 21:22Guido Appenzeller?

  You know, the rebase only carries you so far.

• 21:24Speaker 01

  Right, yeah.

• 21:25Guido Appenzeller?

  You just get too many collisions.

• 21:27Speaker 01

  And they all need a shared memory on the repo they're working on. Because they don't want to reinstall the dependencies every time.

• 21:33Guido Appenzeller?

  That's right, yeah, exactly. And so, you know, you probably need something that's much, much more flexible and more realtime. And, you know, I think we were still figuring out what that is. Realizes doing this is amazing. And I think it's not just, it's not just true for GitHub. I mean, GitHub is one of the big platforms we use. But, you know, take, take the specification writing with, you know, say a Confluence or Jira, for example, what you could use there. If you would develop these systems bottoms up with AI in mind, they would probably look very, very different, right? I mean, it's like, you know, should my story tracker have a function that looks at code and updates stories accordingly? Yeah, that would be very natural, right? So I think, you can do, I mean, I think we've seen this for documentation, right? I mean, something like Mintify looks very, very different at the documentation problem

• 22:18Guido Appenzeller?

  than previous generations. I think we're seeing it for testing, we're seeing it for PR reviews. It's, we're rethinking this entire stack, and that's exciting, right?

• 22:26Speaker 01

  That's true. Another behavior change we have seen, which is really interesting, because humans, like developers, were very lazy.

• 22:34Guido Appenzeller?

  Yeah.

• 22:34Speaker 01

  So if we don't have to read it, we do not read it. We only read the relevant parts in documentation. We only skim through things. So the net new behavior we see all the time with, you know, documentation, like hosted on Mintify is that users who are net new just go in and ask questions.

• 22:53Guido Appenzeller?

  Yeah.

• 22:54Speaker 01

  And agents do not just ingest it anymore. They will actually just do a query against this context.

• 23:01Guido Appenzeller?

  Yeah.

• 23:01Speaker 01

  So context engineering for both humans, because our brains are like LLMs, we need the context, and agents who need context, which is a very critical part. It's such a critical part of like development from here on. And then the question becomes, what are the other tools you think agents will need? Like, we had this huge, you know, market map. From the blog and then there's contrary to where you know all the previous market map, the just which is like here are the developer tools, now there are agent tools to make it better.

• 23:35Guido Appenzeller?

  That's right. In some cases, the same tool caters to both.

• 23:38Speaker 01

  Yes.

• 23:39Guido Appenzeller?

  I mean, that's immediate documentation for agents and for humans. So, look, early days, but I think we're seeing a couple of big categories emerge, right? I mean, sandboxes are helpful to try out, you know, code snippets or build things.

• 23:52Speaker 01

  How do you define sandbox? I think we can record the whole episode on that, but.

• 23:56Guido Appenzeller?

  And I think we will at some point, but the, look, at the end of the thing, you need an environment with certain safety guarantees, right? Where LLMs hallucinate, LLMs can be the clever, you know, if they use external sources, they can potentially be maliciously prompted to do bad things and you just want to have something that basically limits how much the blast radius if anything bad happens. I think we're seeing interesting search tools and parsing tools, like something like Sourcegraph or so, right? If I'm writing a couple of my code in a couple of files, we don't need that, but if they have a really large code base, you know, and suddenly the question is, look, we're looking for, we're trying to replace, we're trying to add a parameter to a function in a library that's widely used.

• 24:41Guido Appenzeller?

  Where is this library used? This is a really hard problem, suddenly, right? I mean, it's like if you're in Python or so, you can import something as something else, so like a simple find operation will no longer find these things. You probably need some tactic parsing to find those. So I think we're seeing documentation tools that are optimized for agents and allow agents to look up these things. I think it's good if an agent can do things like web search, right? So, you know, we're seeing companies in that space. What am I missing here? There's so many more.

• 25:09Speaker 01

  I think there's the other part of more, we're seeing more specialized models.

• 25:13Guido Appenzeller?

  Yeah, I mean, the, for things like code editing or, you know, re-ranking files and things like that, right? That's, that's definitely shaping up as a market. If you take a big step back, if, if we assume. There's massive value creation here that probably creates an opportunity to create a very large number of startups. I mean, if you would have asked me 18 months back, I would have, actually, you had to ask me 24 months back, I would have said like, look, DevTools, it's a smaller kind of market. How big can this be, right? That's not very exciting. If you ask me today, it suddenly looks like this is, you know, a market that could go in the hundreds of billions of dollars, right? In theory, it could go to a trillion, I don't know. So how many companies can you create in that space? I have no idea, but probably hundreds.

• 25:54Guido Appenzeller?

  And over time, it'll, you know, consolidate, but I expect this to be an ecosystem, not a, not a business model.

• 26:02Speaker 01

  I have a fun question, which is going forward. So when GitHub came out, they have this commit charts. It's, it was on T-shirts, you know, like people compare their commit charts, people like commit specific messages, so the commit charts look like, like has a good graphics. What? Because before commits are so tied to the value developers bring, like how many commits do you make? How many lines of code do you change? I mean, we all know it's not the best proxy for value.

• 26:36Guido Appenzeller?

  Yes, not at all.

• 26:37Speaker 01

  Right, so what would be the next commit chart on GitHub? What would that look like? What's the next?

• 26:47Guido Appenzeller?

  That's a amazing question. Maybe how many tokens you burn? If you come to the office, like, look, I burned like 10 million tokens over the weekend.

• 26:54Speaker 01

  But tokens burn could also be very ineffective prompting or context engineering.

• 27:00Guido Appenzeller?

  Yeah, I just stuck my entire code base in the context window, maybe.

• 27:04Speaker 01

  Is it the number of agents you use? Is it?

• 27:07Guido Appenzeller?

  It's even more gameable.

• 27:09Speaker 01

  What is the unit of value that's the closest approximation to what you've delivered as value as developer?

• 27:17Guido Appenzeller?

  Non-cached input tokens?

• 27:20Guido Appenzeller?

  Might be too complicated, I don't know. Yeah. I'm taking this to a high level. The metrics, how we evaluate, suffered about matter changing, right? Where potentially a big, complex refactoring isn't that much work anymore.

• 27:34Speaker 01

  That's true.

• 27:35Guido Appenzeller?

  Because I can let the L M do this and it's structurally easy. A specific optimization in a fairly obscure area where the LLM has no training data, I may have to do by hand and it's vastly more valuable. So yeah, it's complicated.

• 27:47Speaker 01

  Maybe it's number of apps vibe coded.

• 27:52Speaker 01

  I think something else, which is there's on the market map agent toolboxes, there's actually a box I want to double click on, there's the agent orchestration. You can now use not just one agent, but multiple agents, even different copies of the same agent for them to parallelly do things.

• 28:13Guido Appenzeller?

  Yeah.

• 28:14Speaker 01

  What's the implication of this? What can you do with multiple agents orchestrating them together that you couldn't do before? I mean, we're all very ADHDs developers.

• 28:25Guido Appenzeller?

  I think there's a couple of things. Work faster, obviously, right? But you can also try out multiple approaches in parallel and see which one works best. I've seen approaches where people, for example, they want to optimize a certain code. They fire off a couple of agents with slightly different approaches and then just measure which one works best.

• 28:39Speaker 01

  I see.

• 28:40Guido Appenzeller?

  And there's some startups proposing doing that in a more automated way even, right? Where you would just do this, you know, without human intervention. You just say optimize this and then, you know, this gets kicked off in the background. And this all, this all takes a crazy amount of tokens at the end, right? So, I mean, by the way, I think this is a really, another really interesting trend where three months ago, I don't recall anybody talking about the cost of coding assistance.

• 29:07Speaker 01

  Right, yeah.

• 29:08Guido Appenzeller?

  I mean, three months ago, honestly, nothing, right? You go to a Reddit forum on Cursor, so there was pretty much nothing posted there. Today, that's one of the number one topics in those forums, maybe the number one topic, right? Because I think we figured out how with high-powered reasoning models and very large context windows, we can have single tasks that suddenly cost dollars. Not sure about tens of dollars, but at least, you know, dollars is very, very doable. And that adds up, right? It depends, you know, if you're a super high-end programmer, maybe it doesn't matter. If anybody else, a couple of dollars an hour, you know, that's, this is substantial expense. If you're in a low-cost location, it may end up costing more than what you're making. It used to be.

• 29:55Guido Appenzeller?

  That if I was writing software, oversimplifying slightly, had one expense, which is right, okay, they needed, you know, a laptop and some connectivity in an office. But but in the grand scheme of things, the cost that really mattered, you know, at least in the, you know, more high, high cost locations was the the compensation of the person. That seems to have changed now. We suddenly have infrastructure costs for a software engineer, right? They need the the constant feed of of LM tokens to keep them happy. Otherwise, they're not productive. They'll probably change the industry somehow, right? And I think unless we understand how, right?

• 30:31Speaker 01

  We know people will be building more, that's for sure.

• 30:33Guido Appenzeller?

  That's right.

• 30:34Speaker 01

  And the question becomes, does building more correlate with more tokens burned? To some extent, right? And I met so many great engineers who are burning, you know, like they're the top token burning engineer of the company and they're just so effective.

• 30:48Guido Appenzeller?

  Yeah.

• 30:49Speaker 01

  Have like two laptops side by side and, you know, around coding agents that way.

• 30:53Guido Appenzeller?

  It's like the shift from digging versus driving the the excavator.

• 30:57Speaker 01

  Right, it's like rest for research instead of that.

• 31:00Guido Appenzeller?

  But that changes the industry. I mean, probably the person is slightly happier, right? I mean, that's what happened in that case. You need less of them, people will, but but you can also build a lot more, right? So I think it'll change the industry then.

• 31:11Speaker 01

  I think there is more customization to software, to your point about building more. Because there's always one bespoke tool for any business out there. You know, there's HR software, it covers eighty percent of what every company needs. The twenty percent, like, I know this really well because I used to be a PM on these enterprise softwares, we build APIs, so internal teams build their version.

• 31:36Guido Appenzeller?

  That's right, yeah.

• 31:37Speaker 01

  And then now we just have turtles all the way down, like we build the base layer, the internal team builds next layer, and then developers are like, ah, still doesn't work, let me build something else. But now with Vite coding, I actually think customization is just easier and easier than ever. You may or may not even need a centralized team to build that layer using, you know, a commercial solutions APIs. You can just code it up yourself, like what I did. Something I've been thinking about is what's the next workflow or automation gonna be? Like before, we have Zapier and other great RPA tools of the world to make it work. Now with vibe coding, obviously, you still need to know code to some extent to make it work. How would that change?

• 32:20Speaker 01

  I think we'll just end up having more workflows running somewhere.

• 32:27Speaker 01

  The question is like, how can we unlock the net new? Like new developers who are not traditionally technical, but now are writing code to implement these. Maybe they don't need a graphical interface. Or if they do need a graphical interface, it can be represented by JSON, which is more agent friendly.

• 32:45Guido Appenzeller?

  Yeah, in fact, we're starting to see almost self-extending software, right? Software where a user with a prompt can add additional functionality.

• 32:55Speaker 01

  Yeah, that's so true. Yeah, yeah, yeah.

• 32:57Guido Appenzeller?

  Which is crazy. Is that a trend?

• 33:00Speaker 01

  I think it is a trend.

• 33:01Guido Appenzeller?

  Will the next version of Microsoft Word, not the next version, but maybe a couple of versions down the road, have a add feature button in the help menu?

• 33:08Speaker 01

  So software, I think the takeaway point is software is having more affordance than before because of the ability to integrate LM. And what I mean by that is before, if I'm a marketing company, I ship a feature so people can visualize six charts. Now I ship a chat session with LM. LM can reach back to my data and LM's generate code to materialize whatever charts people want to see. So it's more than six. It's like thousands and hundreds and thousands of things that people will want to see. So the interaction model between end user of the marketing software and LM becomes, it can materialize net new features using their own words, so prompt,

• 33:54Speaker 01

  which is very different from before. Software teams are shipping feature by feature. So now with all that, what do you think people want to build? Or what do you think developers should be building? What do you think the world needs?

• 34:08Guido Appenzeller?

  It needs so much. I mean, there's two things I'm sure about. One is, this is...

• 34:15Guido Appenzeller?

  Let's say over the last three, four decades, probably the best moment in time to start a company in the development space, right? It's if you have such a massive disruption, this is what this is what allows a startup to really grow and and and scale and pick a bigger battle with the with the incumbents. I think we've seen that with your GitHub Copilot for Microsoft, first in the market. Relationship to the number one model company with OpenAI, you know, they have the number one source code repo. They have the number one IDE, they have the best enterprise sales force and still, you know, we're seeing, you know, a swarm of competitors that are all doing very, very well against them. So this is this is really the time. The second thing that I'm hundred percent convinced of is that the good ideas are not coming from the VCs, but from the entrepreneurs.

• 34:58Speaker 01

  That's so true, yeah.

• 34:59Guido Appenzeller?

  So, you know, if you spot an opportunity to do something better with AI right now, you can probably add value, right? And then it's about fast executions, about building a great team. You know, it's it's about running very, very fast. But, you know, my prediction is I think we will we will fund dozens of companies in this space going forward.

• 35:24Speaker 01

  Yeah, we are excited to just fund the next wave of startup. But if you're looking for ideas, here are two general directions. The first one is what are the traditional workflows that you can now reinvent? It might not be one, two, one. So like a better Git may not be Git exactly, it might be Git and something else. Right? And then if you just map out the value chain, like what we have on the blog post, you can pick a box and decide what you want to do with it.

• 35:50Guido Appenzeller?

  I think that's right, yeah.

• 35:50Speaker 01

  That's one way to do it. The other way to do it is very differently from before, like as product people, we used to only build for humans, other developers. Now we actually build a lot for the agents. Agents are the customers. Does the agent need a better context?

• 36:08Guido Appenzeller?

  That's right.

• 36:08Speaker 01

  Well, you should build for that. Does the agent want, you know, lower latency for certain models? Well, there are companies shipping, you know, code apply models that, you know, operates way faster with higher accuracy. That's also a need. And when you're looking to where agents don't yet work, there's just plenty of things to work on. You know, from just like easily reusable sandbox. I know there are great companies already building that. Or from to like, how do you enable the agent to kind of smash PR review process with the development process? There's just so much more there to reinvent how agents could work. So tree agents as your customer.

• 36:46Guido Appenzeller?

  Yes.

• 36:46Speaker 01

  And build for them.

• 36:47Guido Appenzeller?

  The classic infrastructure, right?

• 36:51Guido Appenzeller?

  Absolutely. No, I think, I think this is really an amazing time to start a company in this place.

• 37:01Erik Torenberg?

  Thanks for listening to this episode of the A16Z Podcast. If you liked this episode, be sure to like, comment, subscribe, leave us a rating or review and share it with your friends and family. For more episodes, go to YouTube, Apple Podcasts and Spotify. Follow us on X at A16Z and subscribe to our Substack at A16Z.substack.com. Thanks again for listening and I'll see you in the next episode.

• 37:26Erik Torenberg?

  As a reminder, the content here is for informational purposes only, should not be taken as legal, business, tax or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund. Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see A16Z.com/disclosures.