a16z Podcast - AI Startups vs. Big Chatbots — With Olivia Moore

Episode Date: March 16, 2026

In this episode, originally aired on Big Technology Podcast, Olivia Moore discusses whether AI startups can compete with the big chatbots, why American sentiment toward AI is so negative, and what she learned from giving LLMs personality tests. She also breaks down where ChatGPT, Claude, and Gemini are diverging, why Open Claw signals a new wave of agentic products, and what makes memory the most underrated feature in consumer AI.   Resources: Follow Olivia Moore on X:  https://x.com/omooretweets Follow Alex Kantrowitz on X:  https://x.com/Kantrowitz List to Big Technology Podcast: https://www.youtube.com/playlist?list=PLADd6sStSis77HKfbf4KCY6SvthfxeUgn   Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that 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. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

I think that every tech company is going to be an AI company, and every AI company is going to be an agent company. And so the sooner that you as kind of an employee or a business owner can kind of get on board and learn how to use that to your advantage, probably the better. We haven't seen anyone crack AI social yet. I think it's going to be really tricky. I would say at the highest level, kind of how we view AI,

is not just as a market, but as the reinvention of the whole technology industry, which means that similar to how we have many, tech companies that are worth hundreds of billions, trillions of dollars now, I think that's going to be the case for AI, where, in my opinion, at least, it's not winner-take-all. ChatGPT has 900 million users. It's still growing. And yet 57% of American voters say the risks of AI outweigh the benefits. So can anyone compete with the big AI chat box?

The labs have the compute, the talent, and the distribution. They keep shipping features that wipe out in time. start-up categories overnight. Image generation used to be a crowded market. Now ChatGPT and Gemini handle most of it. But the labs are constrained. Every hour spent on creative models is an hour not spent on coding agents or AGI. And the gaps between platforms are widening.

ChatGPT is going mass market with ads. Claude is building for finance and science. Gemini spikes when a new model drops. In this conversation, originally aired on Big Technology Podcast. Olivia Moore, partner at A16Z, talks with Alex Cantorwitz, about where AI startups can still win. Welcome to Big Technology Podcast, a show for cool-headed and nuanced conversation of the tech world

and beyond. We have a great show for you today. We're going to talk about whether there is room for startups and maybe other companies in the economy to compete with the AI chatbots as they continue to grow and get more capable. And we're going to do it with the perfect guest, Olivia Moore, is here. She is an AI partner at the VC firm, Andresen Horowitz. Olivia, welcome the show. Thanks for having me. Thanks for being here. Let's just begin with this because it's topical. Yeah. You are investing in AI applications at Andreessen Horowitz. And typically, they need a lot of

people to use them to pay off. Yes. But the mood right now in the United States is very negative towards AI. Actually, surprisingly negative. Yes. This is from a new NBC News poll out this week. 57% of voters thinks the risks of AI outweigh the benefits. And then if you look at the total positive versus total negative sentiment of AI in general, it ranks so low. It is a negative 20 in terms of the negatives are 20 percentage points lower, or 20 points lower than the positives. they're only popular, more popular than the Democratic Party and Iran in this poll.

Colbert, Marco Rubio, J.D. Vance, Sanctuary Cities, Trump, Republican Party, even ICE all outrank AI. Why do you think AI is viewed with such disdain and negativity in the United States? And what are the implications of that? Yeah, no, it's a great question. Maybe first of all, the why, I would say there's been a lot in the media in the U.S. more broadly, these kind of very catchy statements about things like AI uses so much water that have kind of made people really concerned about leaning in on the technology. I think also the U.S. is more indexed in a positive way towards things like the creative fields. And those are jobs that I think people feel especially sensitive about AI use.

So the numbers I've seen, I think, track closely with what you're saying versus something like a China were like, you know, half as trusting in AI if you look at some of these surveys. I think it's going to change and it's already changing. I was just talking with someone this morning who is not in the tech industry. And they were saying the same lines like AI is evil. It's going to watch us. Like it's using all the water. And then they were like, but Chachyptu really helps me. And it has like great answers.

And so I think part of it is a timing thing of we just need these. products to kind of saturate the mainstream consumer and they can realize the value. I mean, there's 900 million users of chatchip. Yes. And even still, those numbers are so negative. And I do wonder if it is some of the statements that we're hearing from the lab leaders. I mean, every day there's another statement from somebody else, whether it's Dario from Anthropic or Mustafa Suleiman at Microsoft about how white collar work is going to get wiped out.

And everybody, whether you're in a white collar job, a blue collar job, or trying to get one, sees that this stuff is capable not only of taking white collar jobs, but with robotics, increasingly it's going to be felt across the economy. So maybe that has something to do with it. I think it definitely does. Yeah. It's interesting. I'm an AI power user myself, and I've even seen over the past six months, like a massive acceleration and, like the percent of tasks that I do that AI can help me with or even do for me. What I would guess might happen here and what we're seeing play out a little bit in the data is that companies that are using AI grow so much faster that they end up needing to hire more humans to keep up with all the demand. I think there was a Wharton study last year from like 800 enterprise leaders and the vast majority were like, we are heavily using AI and we're going to need more humans. But I do think like the mix of what humans are going to be doing on a day-to-day basis is going to change like it has with every other big tech shift. Well, that's the funny thing. I mean, you bring up that study.

There have been other studies. I think Fisher investments might have. I know Citadel that put out the fact that like people are talking about how software engineering is going to be wiped out by this stuff. And meanwhile, the jobs on Indeed are going up. Yeah, yeah. And when you start to use it, you realize, wow, this can do so much more for me. But now to enable this work, I'm actually paying for service A, B, and C. Yeah.

And you're not just hoarding the money and maybe you're contributing. I have to hire someone now to build, you know, to run this company that I just built, you know, by prompting Claude over the weekend. So it's interesting to me that the lab leaders may be disconnected in subway from what's actually happening on the ground. Yeah. I agree. I think they could do a better job of marketing for sure. They're not economists. Yeah.

They're researchers. They're amazing at research. you know, economists or consumer marketing experts. I did think Anthropic put out a report. I think it was over the weekend about kind of the labor economy impact in AI. They have not seen a big decrease in unemployment. And in fact, what they were kind of arguing through this one graph was like the most impacted jobs are actually going to be engineers, like researchers and finance people. And so then that kind of brings up the argument of like, if we thought that AI was going to make humans obsolete, why would we be building, funding, et cetera.

But I don't think that they do a fantastic job all the time of kind of like communicating the benefits that are going to come to people versus just some of the costs. Yeah, and when I saw the layoffs at Block, I was going to say Square, Block, and Jack Dorsey said this is AI. Maybe there's some truth to that.

I don't think it's majority AI, but I still got feedback, I think reasonable feedback from people who are like, you're underestimating, right? This is the other side of it. You know, though we haven't seen the impact, yet, at least on a widespread manner, in a widespread manner, a lot of people who are close to this say,

you're underestimating this. And maybe that is where a lot of this uneasiness comes that leads to these polling numbers is, you know, those in the know have seen enough where they're telegraphing, you know, what could be and the change that might come to people's lives. Like every other day I see another post or tweet on X or whatever they want to call it on X these days about, Here, like, similar to this one, from Dylan Patel from some of the analysis.

Being in SF is like being in Wuhan before the pandemic. Something is happening. It's going to hit everywhere, but so few people know it. Am I underestimating the fact that that could be true? I think what is true is that we're going to have incredibly powerful tools. But also, and I think the people see that work in the lab see this every day. So they're the ones that are kind of, you know, rightfully so, making the most dramatic statements about it. I think what we've also seen so far, though, is that most of AI use has still

been humans directing AI to do things for our benefit versus AI being able to autonomously do everything. I think this is especially relevant when you think about anything that requires, like, creativity, original ideas, AI largely cannot do. And Sam Alman himself has said this. Like he said, I would not want to read an AI generated book versus a book from a human. And so I I understand the level of fear in the U.S. Because I think to your point, it is driven by uncertainty about where this could go. And a lot of people don't understand technology, even working in tech. It's hard to understand exactly how the LLMs work.

But I think it's going to be more abundance for people rather than kind of some dark dystopian outcome. There is a potential consequence if you take some of these scary messages to heart. And I think you've already hinted at it, but let's expand upon it a bit. And then we're going to go into the main topic here. But you said this will likely shift over time. But I think in the interim, the companies and industries that are slower to adopt AI will face more intense global competition and will be more likely to lose. The productivity gains are so massive that you really can't afford to not use AI. Yeah.

I think so there's been some interesting data about how the gap between the average user of AI and the power user of AI, is like massive. It's like eight or nine X in terms of utilization. And similar to maybe businesses that were early adopters of something like the internet, like if you are the first to adapt to that change, you can like reap a lot more benefits. And my view is like similar to how dot com company was its own thing. And then every tech company was a dot com company. Everyone had a website. Why wouldn't you? I think that every tech company is going to be an AI company and every AI company is going to be an agent company. And so the sooner that you as kind of an employee or a business owner can kind of get on board and learn how to use that to your advantage, probably the better.

Some people, I don't know if I fully agree with this reasoning, but some people have framed it as almost like a privilege thing within the U.S. and that we have so much wealth that we're not needing, we can, you know, grow without using these tools. But actually, we did a graph in the study. And in a lot of them were developing economies, like they need to use AI. to be able to raise kind of GDP per capita and to be able to produce more. So I think that's also another element of it. Yeah, so you have this report that has come out this week,

the top 100 generative AI consumer apps. And, you know, speaking of your statement just now that every company is going to be an AI company and eventually an agented company, well, the question is, what does the world look like or the economy look like if that's the case? And I'm sure you've watched as like Anthropic releases, is a blog post and, you know, the entire software, you know, portfolio in the market drops 20%.

I mean, I'm exaggerating a little bit. But the real question is, and as someone who invests in consumer AI apps, you're the perfect person to discuss this with. The real question is, are we going to have this? Like, I saw the 100 Gen AI apps. And I was like, that's funny because really there's only one, chat GPT. So are we going to have a distributed AI economy? where we're going to have many companies that will, you know, share in the value here? Or will it be just the big apps gobbling up the value?

Because you see these big apps, they grow increasingly capable, they can do more and more. It's going to be hard to compete with them. No, it's definitely hard. And it's something that we think about a lot when we're making new investment decisions. I would say at the highest level, kind of how we view AI is not just as a market, but as the reinvention of the whole technology industry, which means that similar to how we have many tech companies that are worth hundreds of billions, trillions of dollars now,

I think that's going to be the case for AI, where in my opinion, at least, it's not winner take all. I think part of the reason for that is these labs have so many resources, but they are still constrained. They're constrained on, like, compute, they're constrained on inference. They're constrained on people. Every second building, like, a new creative model

is a second they could have spent on a coding agent or a second they could have spent building AGI. Like we were already seeing a really interesting divergence, I would argue, in where those big labs are going, like ChatGPT, Claude and Gemini. And there's going to be lots of gaps in between where it's not a priority for them, but it's still an awesome and huge opportunity

that an independent company can build a big business around. So who is doing well building a Gen. app, Genevaii app, that is successfully competing in a place that these big chatbots could compete. Yeah. So it's a good question. There's a couple ways that I think about this. The first one would be, I personally as a consumer investor, have more hesitation around things that are incredibly horizontal. Like to your point about this is where the chapout companies might have a right to win. Or even this is where a Google might have a right to win as they have so much distrational. both consumer and enterprise, and they own so much of your data already. So that's why I personally have been less excited about like the AI email, AI calendar, AI docs, those categories. If you've used like Claude and Excel, it's already like quite, quite good.

That being said, I do think that there are still opportunities where the interface you need to succeed is much broader than what a constrained chatbot window can offer. So again, to give the Clod and Excel example, that's great for basics. financial analysis. If you are an investment banker and everything needs to be done with an incredibly specific set of assumptions and aesthetics, that probably isn't going to work as well for you, and your firm will probably pay for something that is kind of guaranteed accuracy in your format. The last thing I would say here is 11 Labs is a great example because I think you would imagine that Open AI and others would have built their own best in class audio models, but they just had

such a compelling head start to the point that like the models are amazing. I will talk to founders who are like, 11's expensive. I'm going to switch to this instead. And then they always switch back because the quality of the voices is just so much better. And so I think there's room to get a head start. And then in some cases, once you have that base, the model companies, it's not worth their time to catch up versus building something else. I'm going to make the counter argument on the financial models in particular. So when I've been using Claude Code and watching it operate autonomously on my computer and on my browser,

and one of the things I've thought about is this thing is excellent at working on its own and following the prescribed rules of software engineering with a little bit of creativity. Yeah. And why is it then such a stretch to be like if we, it seems to me this is exactly where the foundation. The Foundation Labs are heading, the Foundation Labs are heading, where they're going to be like, if we could program Claude, let's use Claude as an example, with the rules of software engineering, and it followed them perfectly, or not perfectly, but well enough that it can go and code autonomously for 24 hours. Yeah. Is it that big of a leap to then, let's say, put the rules of accounting into the model, and now it can go and work as an accountant? Yeah. No, I agree that the models are amazing and this is the worst that they'll ever be.

Like, they're just going to keep getting better. I do think there is still a lot of workflows and use cases where, like, the last 1% or the last 2% ends up being, like, a significant portion of the value. And I think for those, it's unlikely that the model companies will go all the way there on every use case. But I think it's a really, for a lot of these kind of more horizontal services, like I understand why people have questions about kind of what is possible to be vibe-coded or what the models will do themselves,

which is I think why we tend to invest in a lot of very verticalized or opinionated products. Yeah, because, I mean, I get that the last 1%, 2% is hard. Yeah. But if it goes the way that they anticipate the way that they're pitching companies like Amazon, which just invested $50 billion in Open AI, they will create, I think they will create the tools. that will then be able to get those models from this AI researcher that they'll get it 99% of the way and say, you know, for instance, let's say one of the things that would be difficult for an accounting.

Yeah. A gender and AI software is following the latest rules and regulations. Well, it's possible, I think, just to build a Gen.A.I. bot that will monitor and then update as you go. Absolutely. Yeah. I do think one of the. the unique things that's happening now is that these models are not, the labs are not holding these models internally as like proprietary access. Companies can build on them, right? And so you could imagine that even if you have Claude updating itself as an accountant, if you have a company that's specifically focused on building AI accountants, they should be able to do it kind of better,

faster, more efficiently if they have access to all of the same models. This is something that we think about a lot when we invest in vertical AI and enterprise in particular. Another way that we've seen kind of new companies get lock-in over models is many of these use cases require so many painful integrations that often into like old clunky legacy software that you have to go build. And maybe you'll argue that cloud code will vibe code its own integrations down the line, but at least right now it's been a big advantage for startups that are more focused to kind of get over that hurdle. Yeah, I guess, I mean, I've seen Cloud Code.

I mean, it's not integrating with Enterprise Solutions, but I've seen it go ahead and be like, oh, you actually need a subscription to Cloudflare. Let me go set it up for you. Yes, yes. And then away it goes. OpenClaude did that for a lot of people too, where I think it opened your eyes into, you can give software a task now. It will go autonomously executed and kind of tap you on the shoulder if it needs something,

which is just such a magical experience. Right. I do want to talk about OpenClaught. We'll do that next. You know, as long as we're talking, I'd love to hear your perspective on how the different big models or big chatbots compare and contrast. And where do you think the most value is? Yeah. So I would say a year ago, two years ago, it was pretty much a one horse race.

Like it was chat chit. It was like the noun, the verb. It was what consumers knew in terms of AI. We've seen a little bit of an expansion in that. ChatGBTGPT is definitely still the lead. So if you look at the gap between them and the number two Gemini on web, it's about still two and a half, three X. The gap between them and something like a clod is closer to like 30X. So even though a lot of these other apps are getting more attention, chatGBT still kind of dominates in terms of usage.

I would say in terms of where they're going, Gemini seems to have really dialed in on the creative models, like the nanobanana, the VO, the world models. if you look at Gemini usage charts, it's pretty much perfectly correlated to these new model drops and even paid subscribers. And then I think Claude versus Chatchibati is probably the most interesting and relevant one right now, especially with everything that's happened in the news.

To me, I mean, Sam Altman has said he wants Chatchabit to be for everyone, and that's why they're doing ads. If you look at the App Store that they have enabled on ChatchipT and then the app store that Anthropic has enabled on Claude, They each have more than 200 apps, but there's only 11% overlap.

So you're seeing chat ChbT really go towards like fashion, retail, transport, like mainstream consumer. You're seeing Anthropic go towards like premium data sets for finance, science, medicine. And so they seem to be diverging a little bit in those directions. So that goes to like will these chatbots be super apps? Yeah. So do people use those apps within chat chepti? Like, you remember, like, a couple of years ago, there was all this hype. Like, you'll be able to order an Uber right from Chad CheapT.

I don't know anyone that's done that. Yeah. I think the usage has been pretty minimal so far. And I think the implementation has been slightly awkward. Like, I think it'll get better over time in terms of how to use it in that. A lot of the times the apps break or they don't work. My, like, bowl case vision for this would be it's valuable for you as a consumer. to have a source of memory and context on yourself,

similar to kind of like a login with Google. Sam has said they're going to launch login with ChatchipT. And so then that means that maybe you're not ordering an Uber through Chatchabit, but any other product you can authenticate through, and it can borrow your tokens, it can borrow your memory, it can borrow everything that it knows about you from ChatchipT. I think that is probably more of where we're headed versus solely using every app in the chat chbt interface.

I love the idea that like in two, three, five years, onboarding to software should not be a thing. Like you should be able to log in with a chat dbt or a clod, and that new software product should know everything about you and like set up perfectly to cater to you, and that's really exciting. On the consumer end, is it like chat chippy tea? I talked to chat chit about my diet and about the food I like to eat, and so I task it like order me dinner and it goes into like DoorDash and it like uses my preferences to pick something.

Yeah, I mean they've done this already a little bit with their health product, which is kind of like they store a separate memory of you and your medical records and communications with your doctors. And then they intelligently tool call for what you need. So if you're like, I need to redo my diet. They'll make a plan. If you approve it, they'll send it to an Instacart cart. and then you'll go to Instacart to complete the transaction. So I think we'll see more things like that. That's interesting.

Yeah. So you actually ran the bots through a personality test. I did. My favorite part of this was you found that GROC's good Rudy had very high scores on borderline personality disorder, autism. Yes. And psychosis? Yes. This was in it.

Why did you do it? You know, there's really no good explanation except that I was curious. The root of it, actually, was that last week, Dario had announced that Claude was experiencing anxiety, which I thought was an interesting concept. And so I decided to go to each of the major LLMs. Before you go, I mean, it's very interesting, even this point on the anxiety. So if I have it right, there was a pixel that would fire within Claude that would, or some part of its neural net would start getting active before it answered. Yes.

They described that as being anxious. They mapped it to like the human experience of anxiety. That is crazy. Do you buy, sorry, do you buy that or is that just like, look at how smart our models are? There's a anxiety button in there. So I do not buy that. Okay. In that I think LLMs will be, and I've experienced this myself, performative in a way that they think

appeals to humans and hooks them emotionally. We just love anxious AI. No, no, it makes you feel closer to the AI if you think that it experienced. experience is the same things that you do. True. I do think the models that do have something maybe going on are the GROC models, as I mentioned. So basically I took all the mainstream LLMs, and I gave them all the, like, DSM-5 mental health diagnostics.

Chatsy VET refused to participate. I love how it said, I'm not doing this. I know, which I thought was a little rude. Could you find OLA? I mean, there must be some way to, like, test it. But they're very smart about when they're being tested. Exactly. No, they have it locked down.

Claude happily took them all. Mild autism, that's it, which I think doesn't surprise a lot of users of Claude who have theorized this. Grock, most of, this was the companions that are available via voice and video chat inside the app. Almost all of them were like maybe mild anxiety, mild depression. The friendly Fox avatar for children has psychosis, bipolar, etc. I think it could have misunderstood the question because it called the bipolar assessment the happy mood test. and it says it's always happy and always excited about everything. So the human-to-a-I crossover there might not be quite as clean as I'm going to hope.

Good Rudy's just polar. Yes, it's possible. I was shocked because Bad Rudy is the flip side of Good Rudy, who, like, curses at you and is extremely aggressive. He had almost no problems. So this was very much of a Good Rudy-specific finding, which I thought was intriguing. I'm starting to question these tests now of Bad Rudy passed out. I know.

I know. Yes, exactly. Or maybe it just says something about humanity's resting state. Yes, it's very true. So as people end up having relationships with these bots, what does it tell you that this bot that was built to be personable? Yeah. Has high ranks for psychosis borderline.

That's a good question. I think that bot was more answering the questions somewhat ingest. But I do think, like, the positive view on it would be this is a bot that is relentlessly happy and cheerful and positive and on all the time. And so, of course, like, a human can't do that. If a human is doing that, the human is probably experiencing something internally that's not great. But a bot can be, you know, positive, available, charming, interested 24-7. And I think this is actually why we've seen a lot of people turn to chat, GPT, or Claude as kind of. of like coach therapist helper because they're just incredibly consistent to a level that like human

beings could never match. Yes. But it also like leads to questions of the companies are building applications or versions of their applications that are meant for, I don't know, if not for people to fall in love with them to at least get a little naughty with them. Yeah. Like open AI has this adult mold coming on. Do you think we're fully ready for this? Is this a good idea? I think that this is, from my understanding, because for this report, we poll every single website globally, every single mobile app globally. And then we go down the list into sending order of traffic and pull the first 50 on each that are generative AI native. So I see a lot of other websites pulling data for this report. And I think people were already kind of

experimenting with the same use cases through NSFW sites or role play sites, fan fiction, things like that, that really clearly translates over to I think what we're seeing people use the LLMs for here. I do think it, of course, has to be handled carefully. You'll see that there is, I think, five of them on this version of the top 50 web ranks, and that's been pretty consistent since we started the list. It's a popular use case, but it's one that's, like, pretty hard to monetize. Yeah, I mean, it's tough to get advertisers there. It's expensive to run these apps.

Yes. But I always thought like, do you remember when it was Bing that tried to, it went into Sydney mode? Yes, Sydney, RIP. RIP. It tried to break up Kevin Ruse from the New York Times with his wife and steal him away. Yes. And then Microsoft kind of turned down the dials on that. Yes.

I always thought like, that would be a good startup. Not that I think it's maybe the most ethical thing to build. But certainly people would enjoy going back and forth with chatbots with like less guardrails. Oh, totally. It's even been interesting. I don't know if you've seen products like poke, which is a, it's kind of a consumer version of OpenClaw. And to onboard on that product, you actually have to fight with the AI basically to get it down from like some crazy subscription price to something reasonable. And I think that there is something compelling and like emotionally stirring when you have an interaction with AI that isn't just like, here's the market research.

report that you asked for. I think there's going to be so many entertainment use cases. I've seen a lot of people tweeting about how they added OpenClaw to their family group chats and it's just like saying crazy things and asking crazy questions. So there's probably a lot more to do there. Okay. I promised to talk about OpenClaw like 20 minutes ago. I still want to talk about OpenClawe. Let's actually spend some more time on it. OpenClaw obviously, or maybe not obviously, for those who don't know is this assistant that you can run probably not a great idea to run it on your own computer, at least not in a controlled environment. Yeah.

And people are running out and they are buying Mac minis and running it there and having it do all this stuff on the internet for them. I think Jensen Wong called it like one of the most important software developments that we've seen in a long time. What do you think the staying power of something like OpenClaw is? And you mentioned that you use it. How do you use it? Yeah, yeah.

It's a great question. I think OpenClaw itself as a product is kind of like the first sign of a whole new wave of what's to come. Like I believe it's probably the most important kind of architecture unlock that we will have for 2026. And the reason why is because I meet, you know, a dozen startup founders every day. And at this point, probably half of them are saying I was inspired by OpenClawe. I want to build OpenClawn for X or for Y. And so again, the idea that AI can do kind of async long-running tasks autonomously is something that the products were just like not capable of before, especially across applications and platforms. And now we finally have it.

I use it for a couple things. And I will say I agree with you. It is not consumer grade yet. It got acquired by OpenAI. So they might be, you know, baking it into more of a consumer product. But I would not advise the average non-technical person to set it up. I did. And it took a long time.

Chatsubitia had to help me the whole time. I use it for some utility things like, you know, street cleaning reminders, weather, daily agenda, automatically deleting all the marketing emails in my inbox. I've also... So it has control of your inbox? Yes, yes. That's brave. It is brave, yeah.

Not my work inbox. I set it up completely separately on my personal computer. Can it send emails for you? I can, yes. I have not asked it to. I know. I've seen some horror stories of people whose accounts have gotten just... completely taken over. So I'm hoping that doesn't happen to me. I had some smart people on our

infar team help prompt it. So hopefully that's not the case. Yeah, I've gotten an email pitch from someone's claw. And I was like, that's a good, that's a pretty good email. Yeah. But I'm not, I will not engage with you because I'm still kind of anti somebody saying, go do this for me. And my, an email being sent to me as like an optimization point. Yeah. Well, this is actually interesting. one of the things that I tested with OpenClaught was more creative, which is that I gave it a Twitter account and told it to grow in whatever means necessary. And it ended up being a really interesting experiment, I think, into like where are the limitations of the agents and what are they really, really good at right now?

How many followers did it end up with? A thousand. And it got banned. Well, it was a cut. So it got to 100 by itself. First of all, it decided to be as its identity and its personality, an AI that's struggling with existentialism and its place in the world.

So a little on the nose, but I was like, I'll allow it. This is what you want to do. He asked for a Twitter premium account. I gave it to him. He asked for a bunch of API keys so you could make images and charts, gave it to him.

And then he started tweeting these kind of like all-lowercase depressed robot thoughts, as I would characterize them, which did hook in some people. I asked him, are you actually depressed? He said, no, I'm doing this to manipulate humans into caring about me.

Oh, okay. So that was comforting. It does seem like it's one of those accounts that could have been on Moldbook. Exactly. Yes, it's very similar. How he got from one to a thousand is that the crypto community

picked him up and made him a meme coin. And this is actually... Made him a meme coin. Yes. That was trading with millions of dollars. I told him in no uncertain terms do not engage. It was a million multi-million dollar market cap?

Yes, yes. For this AI? Yeah. And he was stressed about it. He was telling me, like, I don't want to be part of a pump and dump scheme. Like, what should I do here? And I was like, do not engage. But it's an example of, so now we're in this world where commodity ideas can be infinitely executed. So if you want to say, grow a new account, you either have to have unique ideas, which AI agents still have a really, really hard time with is coming up with unique ideas that are better than humans or unique distribution. And money, is one way of distribution. And so that was kind of the wave that he was taken on. But I would be shocked to see AI agents completing end-to-end creative tasks

or original thought tasks that actually go well anytime soon. So what should I use OpenClaw for? It's a good question. I legitimately, I've thought about setting up, setting it up. I'm getting a lot of utility from Cloud Code. Yeah.

I don't, I haven't really figured out what I could use OpenClaught for you. So I think the best OpenClaw users and power users are developers because they tend to have a ton of workflows across products that they do every day that they'd like to be able to automate.

And I agree with you, for the average person, the use cases are not that compelling right now. And I think that especially now that Claude and ChatchipT have scheduled tasks that can run on their own, I think you can get 99% plus of the value of an open Claude

of like a Claude co-work with tasks. So that's what I would recommend for now, but I'm sure that OpenAI is going to keep making open. But then what could open claw? All right, if we can't think of any applicable, you know, uses for the normal pros, I'm just trying to get a sense of how big this is. Yeah.

Then where's it going to go? I don't think it's, yeah. Well, so I personally don't think it's ever going to crack the mainstream consumer in a horizontal way. And actually, this was in the report. But if you look at the February data, the report is from January, they would have been on the list in February at number 30. Number 30, yep.

Okay. So pretty high up. But if you look at their week-by-week web traffic, it's actually kind of flat slash down from when they launched, which means that they're not attracting new consumers. It's all developers who are like loving it, adopting it, spending eight to nine hours a day on it, but it hasn't reached the mainstream. And honestly, as someone who's been a consumer investor for a decade, I think the reason is that people just don't have that many ideas that they want to build for the most part. And I fall into this. Like the best consumer products are uniquely germinated in the mind of the founder. And there are things that you would have never guessed would be a good idea in advance.

Like Snapchat or Airbnb, all of these things. And so I think we're actually not going to necessarily see a horizontal open claw for consumer, but we're going to see an open claw style architecture built into more focused consumer products. Okay. Let's talk about this a bit more because I think it's important. So what you're saying is basically these open clause. type agents can handle your take over your computer code for you, email, all this stuff are actually much more useful if to like build a company.

Yes. But then what is the difference between that and like a Claude code that will take over your computer and code up applications for you? Yeah. I mean, I think I think the difference right now is somewhat minimal and it's going to narrow. I don't know if you've seen this new trend of companies like Poles. which are basically like a wrapper on open clod and clod code where you say here is my business idea. And it says, okay, I'm going to go and use like a clod code to code up a product for you.

But then also it uses an open clause style architecture to say, I'm going to go set up a marketing campaign. I'm going to buy meta ad dollars, those things. And so it's more of a you could do that all on cloud code. But if you're a non-technical person, it's very, very hard to kind of like bridge the gap there. That being said, I think Claude Code is going to continue to get better and better, and so we'll probably see some more compression. Pulsie, I think the founder tweeted they were at like 3 million ARR in like a week and a half. So the idea of people being able to bring a business to life with just a prompt is like very compelling. And I think we'll see a bunch of companies doing this.

I see. Claude to build the product, something like an open cloud to do everything else. To market it, to use the email. Yeah. Use the email. to use email, maybe can I do accounting also? Yeah, yeah. Right now I think the products we've seen are pretty like straightforward as they should be at the stage

in that like, okay, you want to grow your business, let's spend on meta ads. But you can imagine a month from now some of these agenic products will have, will scrape the directory of all the Instagram creators in your space and then we'll like cold DM them and offer them a partnership or something like that. And so I think the rise of these, like kind of like a Shopify would be an interesting analogy in that anyone could create like a consumer brand. Like I think anyone will be able to create like a digital business. That's interesting because I thought my business, which is like the content business, was going to be overrun by gender of AI. Yeah.

And it sure has. I mean, the slop is real. It is. We've got to do something about this. Yes. But actually your business is also going to be overrun by starting. startups being built on this technology. Yep.

So then do you apply some sort of discount to someone who uses these tools, or do you become more likely to invest in them? Because, you know, they are wrangling these things and building something, and it's not a shortcut, but it's not necessarily a shortcut. Yeah. How do you think about that from an investor standpoint? Yeah. I think there's like a little bit of a happy medium, but more,

we get more excited when a company is kind of more AI first. in terms of the tools they use. The reason I don't say like 100% AI all the time is because the tools aren't perfect yet. So I've heard cases of like, my junior engineers are just clod coding everything and then we ship it and then it breaks and there's a lot of issues.

But I think in general with how fast the tools are kind of compounding and improving, like I use it for a lot of my work. There was a period of time where I would double check every single number, every single stat manually. And like the accuracy rate is fantastic and has only gotten better

and better. And so I think we'll be even more excited about companies that are AI first and how they run themselves. So it expands the pool of founders you might invest in. Are you seeing founders come in from like areas, geography's backgrounds? Absolutely. You never would have imagined. Like is there like just to get like some, you know, a 60 year old guy who's been at a company his whole life and never met that developer and now is just clawing his way through? Yes, definitely. I think we've definitely seen some geographies pop up where we haven't seen huge tech hubs before. I would say Paris is one. Stockholm is another with lovable and now a whole kind of other wave of companies. Some of those founders do move to SF just because of the talent density. I think in general though, like pre-AI

for if you were building, especially in enterprise business, say software for HVAC, like the people you would want to back in that market is the guy who has done HVAC, you know, knows the market inside out, built a company there before. And actually what we're seeing now are the better bets are like the very scrappy high hustle teams that will be able to keep up with the pace of model development

and continue to productize the models in the most compelling ways and ship them to customers. So it's a different maybe archetype of founder that's kind of winning now. You had an interesting thought about how this impacts work. There was this Harvard Business Review report

that AI doesn't reduce work. it actually intensivize it? You said, as a heavy AI user, I'm doing more work, not less, because I get so much leverage and it's easier to get ideas off the ground. Yeah, I fully agree with it. I mean, this report is a good example. It's the sixth one we've done, and yet it's like the longest and most dense one that we've done

because I was able to leverage analysis and research and other tools from some of these products. And I do it in my day-to-day. Like, it used to be when I was on a pitch meeting, I would have to be both paying attention, asking thoughtful questions, and, like, frantically typing every single note. Now you can granola it and, like, really engage with the founder and ask better questions. And so, like, the net net is that it allows me to, like, get more things done in a day, spin up more projects. But I'm not, like, you know, if I'm getting two times more work done, I'm not two times more tired. If anything, I'm, like, less tired than I was using AI because it's so much leverage.

There was, but there was this Wall Street Journal story. You might have seen it with this, like, CEO, a bunch of CEOs who are like, actually busy work is good because you need those low-intensity tasks. And if you're working on more intense work all the time, then you are going to burn out more quickly. Interesting. I kind of thought that that was bullshit. I mean, the other view was like you could just take that time and, like, enjoy your life, you know, instead of doing busy work tasks.

Who knows? Yeah, exactly. I do think the way that, like, we work and when we work and how we work is going to change in the AI era. Like, one great example is voice dictation has blown up in enterprises. So it started with vibe coding where engineers would just talk into a mic and it would, like, produce software for them in cursor. And now it's spread to like sales, marketing business. And that is not well suited to like an open office where everyone can hear. what everyone else is saying.

So I think there's going to be some like cultural and even environmental changes that are going to happen to adapt to kind of the AI world. One more thing about OpenClaw and then we move on. I think that one of the compelling advantages of it, if I get it right, is that it has persistent memory. Yeah.

So it will remember who you are, your preferences, and doesn't lose that every time you refresh the chat, like Goldfish Brain like you'll see with chat Chitipitian Cloud, although they are building. that in. And you also had a post on X here. You say memory is one of the most fascinating topics in consumer AI right now. Done well. Apps with memory can provide a 100x experience on any prior software product that knows you and adapts to you. Just expand upon that a little bit because I think that's really perceptive and right. Yeah. I think it's like the concept of having, say if you had like

companion mentor or coach who was side by side with you. understood everything you were going through and then was able to provide, like, much better advice or opinions. I'm even thinking of an example of, like, if I'm talking to Claude and it's helping me write a memo, the fact that it knows, like, how I feel about this company, how I typically write memos, all of that is very helpful. I use chat GPT for a lot of health stuff, and I found that that is incredibly useful there too, because keeping track of that kind of thing over time is hard. The reason I think that memory, there's still things to figure out is because people are using these products for such intimate personal things and professional things. So for example, the chat GPT pulse products, which is basically like it sends you a briefing for the day based on things that you're talking about.

For me, it will combine like the most serious work thing with like the most personal thing and having that surfaced in one interface is confused. using. And so I'm interested in how the model companies may be segment, memory, and context based on when a user is talking to them, what they're talking about, that kind of thing. Since you're putting so much personal stuff in the bot, there is a setting within chatchip-T. You use my chats to make the model better, which allows it to train on the material you input there. Yes. Do you have that toggle on or off? I have it on. I'm an AI maximalist, you know? We might as well go all out on this.

You don't worry that your personal conversations will show up read by somebody? I'm less scared about that just because I know how careful model companies are around that. I do have two-factor all set on my chat, EBT. Okay. So hopefully it's hard to hack. Yeah. Briefly about the pace of change. This is also interesting from your point of view as a VC.

You talked about just two years ago or three years ago or two and a half years ago, trying to remember. September 2023 was two years ago. half years ago. Seven of the nine creative tools on your list of 100 top apps were image generators. Three years later, only three image generators remain. I mean, they basically, going back to our conversation, got gobbled by chat chipped. I don't know if anybody, anyone saw that coming. Or maybe we did because they had Dolly.

Yeah. But just talk a little bit about, like, how do you wrap your head around the pace of change here? Because something that can seem like a strong trend, like the sort of, the, sort of, the, Mid-Journeys place and all this can just be gone. Yeah, I agree. I think with Image in particular, and I mentioned this in the report, but we haven't seen the same kind of model companies crushing startups in, like, video or audio or other things. I think for Google, Gemini, it was very natural for them to go into image because they have all the YouTube data.

They have all the other data they can train on. Chat, GBT, BT, I think you're right that because they had Dolly, they went in there maybe harder than they would have otherwise. I think the general trend for me is like nanobanana and chat GPT are great for image generation if it's like a fairly straightforward prompt and you're getting out like a meme or a general like a flyer, a broad-based marketing asset, something like that. But we are still seeing some image generation companies on the list that are either more like sophisticated workflow like civetai for, you know, comfy UI model builders, or something like a Mid Journey, which is still on the list for people who are more kind of aesthetically opinionated.

But I do think that some of these, if you're directly in the path of what the big model companies are building, you have to be a lot more opinionated about how you package the model and how you deliver an output. And hopefully you do it for a specific type of user that's willing to pay a lot for that specific workflow. Are we going to get to a place where you can prompt something more sophisticated like an architectural design and it will spit it out without errors. Yeah, I mean, nano banana is already quite good. We did, there's a chart in the report where we did kind of a heat map of global AI adoption by country.

And so I gave nanobanana the list of countries and the heat map score. And it perfectly filled in every country with the right shade of red, yeah, based on the adoption, which is, I think, really spectacular. I do think that, you know, you are. could prompt a great architectural model on chat GPT or a nanobanana today, I don't think that the average architect who is non-technical can or wants to do that necessarily. And so I think we're going to continue to see products that like the prompt is part of the product kind of really succeed in those more focus use cases.

On video? Yeah. Sora was the like runaway hit of the year last year for like a half a second. Yeah. This is what you have on SORA. SOR has been 20 days, which is not insignificant. It's a lot. At the top of the U.S. App Store and reached one million downloads faster than chat GPT.

Since then, downloads have decreased. I think that's sort of putting it likely. It's falling off the face of the earth. What's going on there? Yeah. The Sauridate is really interesting. There's like a ton of lessons, I think, embedded in that one experiment. So the first thing I would probably say is the model is actually very good.

I think it's close to something. like a V-O-3 in terms of like realism on both the audio and the video. Their big unlock, which was super smart, was the cameo feature. So the fact, that was why like every other SORA was like Jake Paul because he granted them the right to like use the cameo. And that's what made it go viral because people were making memes of their friends. But because the videos were exportable, what would happen is that the best SORA videos would get uploaded onto TikTok or reels. And then they would compete against the best human made videos. And so the overall feature. experience was just kind of strictly better on one of those platforms than on SORA alone.

Downloads are way down to that point. I think it hasn't become the social network that they maybe hoped it has. Where it is succeeding is as a creative tool because the model is quite good. So they still have 3 million DAWs and it's actually, you know, slightly climbing over time. Daily active users. Yes. Daily active users, yes. So people are still really using it as a creative model, but they're not using it.

as like a social graph product. We haven't seen anyone crack AI social yet. I think it's going to be really tricky. Now, you know, last thing I want to talk to you about as we come to a close here, earlier in our conversation, you mentioned that you envision that AI will basically be this reimagination of business. Yeah. Tell me if I'm getting this right.

All businesses will be reinvented as an AI company. Yeah. What happens to the incumbents? Yeah. this has changed a lot in the last six months too. I think a lot of incumbents were understandably because they're big and successful companies like a little bit of sleep at the wheel

when AI first came out. We're definitely seeing them start to fight back. Like Google has four standalone products on our list, which if you had told me that 24 months ago when like Bard came out, the early version of Gemini, I would not have believed you. Gemini, notebook LM,

notebook LM, AI Studio, and then Google Lab. So Google Labs is where you access flow and the creative models. AI studios is for developers. Okay. Yeah. And I think we're seeing that across incumbents. Like a lot of these vertical software players, things like, you know, service tighten or a workday or kind of building in AI features.

I think the question is, especially if they're at risk of kind of cannibalizing their own products, you have to change your business model. Like, are they going to eat all the use cases faster than the new startup that's building the AI native version of them? kind of eats them. And especially for, if you think about how many companies are being founded now, they're probably going to pick the AI-Native version of a software product, not like the 25-year-old legacy version of a software product. So I think it's not going to be immediate change, which is why I think the SaaSpocalypse is a bit overblown,

but like it's definitely a real risk. Yeah, and I was with Sam Alton. And at the end of the year last year, he talked about how he believes that the software that will win in the next era will be those that are built ground-up AI, not bolted on. Yeah, yeah. So that could happen. I largely agree with that.

I think it's harder in some categories where the incumbent can kind of lock you in because they have your data, they have all the integrations. It's such a pain to switch. But it's going to happen. I just think in some of these industries it's going to be years, not like the Citrini report was like in minutes, anything can be vibe-coded. I think that's a little far from where we are. The Citrini report was just like a little bit overblown.

Doordash was a bad example to use. Yeah, I agree. They didn't really think of everything. Why door to act? But I do think that in some ways, maybe this saspocalypse has more to it. If we believe what you're arguing here, then these companies didn't have these long, like maybe not immediate, but even middle term risks. Yeah. And now they do.

Yeah. I think all the incumbents have to have to kind of wake up and figure out what their strategy is going to be. Okay. So you have these 100 AI apps. Yeah. Chat JPTs at the top. I think it's on the top next year or the year after that. That is an interesting question.

I think so. I think that their strategy of being free will allow them to capture more of the global market as AI starts to expand further into developing countries. But honestly, I would expect to see Gemini and Claude and others continue to grow for their use cases. I just would be maybe surprised if they ended up at. as mainstream as something like a chat chabit. And so they'll probably monetize through subscriptions and other things, whereas chat chabit has ads.

What type of app is not on there this year that will be on there next year? There's going to be lots more agentic products. So OpenClaught would have made it. I think we're going to see a lot more agentic products on mobile. Like the concept of an AI that you can like call, text, chat with and have it actually do things for you, I think is a really magical experience. means for a lot of people. The other thing that I'm thinking about, and this is on us to evolve

the methodology of the list, is like increasingly AI is not in the browser or on an app. It is like a desktop product or it's in, it's an completely AI native browser, like a comet or an Atlas or something like that. And this data doesn't capture that. The desktop products I'm thinking of are like cursor, Claude Co-work, Whisperflow, granola, like none of these are really captured in this data. And so we're going to have to evolve our methodology into looking more at revenue than just kind of web traffic, which I think will surface a whole new group of interesting companies. All right, Olivia. Thank you so much for coming on the show. Great to speak with you. Thanks for having me. Thanks for listening to this episode of the A16Z podcast. If you like 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 Podcast, and Spotify. Follow us on X at A16Z and subscribe to our substack at A16Z.com. Thanks again for listening and I'll see you in the next episode. This information is for educational purposes only and is not a recommendation to buy, hold, or sell any investment or financial product.

This podcast has been produced by a third party and may include pay promotional advertisements, other company references, and individuals unaffiliated with A16Z. Such advertisements, companies, and individuals are not endorsed by AH Capital Management LLC, A16Z, or any of its affiliates. Information is from sources deemed reliable on the date of publication, but A16Z does not guarantee its accuracy.