A New AI Investment Fund's Strategy and Opportunity — With Techmeme Ride Home

0:00Two investors with backing from Mark Andreessen, Chris Dixon, Dennis Crowley, among others. Join us right after this to talk about their new fund. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge, and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade in value. advanced intuitive and deployed that's how they stack that's technology at capital one the truth is ai security is identity security an ai agent isn't just a piece of code it's a first-class citizen in your digital ecosystem and it needs to be treated like one that's why okta is taking the lead to secure these ai agents the key to unlocking this new layer of protection and identity security fabric organizations need a unified comprehensive approach that protects every identity, human or machine, with consistent policies and oversight.

1:04Don't wait for a security incident to realize your AI agents are a massive blind spot. Learn how Okta's identity security fabric can help you secure the next generation of identities, including your AI agents. Visit Okta.com. That's O-K-T-A dot com. Welcome to Big Technology Podcast, a show for cool-headed, nuanced conversation of the tech world and beyond. We have a bonus show for you today. Yes, a bonus show. Two investors, Brian McCullough, Chris Messina, here with us today. They're starting a new AI fund, and I thought it'd be a perfect opportunity to talk about why you'd start a fund with AI right now.

1:40Are there really moats? Is there anything investable? And for anyone thinking about what the next generation of startups are going to look like, well, you're about to hear them. Brian and Chris, welcome to the show. Thanks. Thanks, Alex. So the first question for you guys is, are you bananas? Because I've seen so much discussion on Twitter about people talking about how there are no moats for AI. Very like prominent investors, Keith Raboi, Sam Lesson saying that they think that anybody investing in early stage AI right now is relatively insane because the value is ultimately going to go to the big tech companies, Amazon, Apple, Facebook, Google, Microsoft, and the chip makers like NVIDIA.

2:18So where exactly are you aiming to play? And why do you actually believe? I mean, you've convinced Marc Andreessen and Chris Dixon and others to join in. So where are you going to play? Everyone we talk to, no matter what the nature of the call is, eventually it devolves into what are you seeing? And like literally the conversation with Marc Andreessen was the same where he's like, OK, guys, I'm in. Now, what are you seeing? Like it is it is. But everybody's it's it's wide open. I mean, like literally, if you if I know that, you know, AI is a thing has been going on forever. But like this moment is is barely six months old, which actually Chris and I can speak to that in terms of why we're doing the fund.

2:57So you're calling us investors. We've been on your show before. I'm a podcaster. Chris invented the hashtag. But essentially investors, but now that you're starting to fund that the label you get. So, OK, OK, let me let me let me explain as fast as I can. So what we're announcing today is the Ride Home AI Fund, which is a$15 million target fund doing seed, pre-seed, Series A in this new AI technology. Most relevant to anyone listening, it's a 506C fund, which is why we can talk about it today, which means any accredited investors listening can invest. So if you're having AI FOMO and want to participate in this moment, but you can't get into an A16Z fund or a Thrive fund, you can get into our fund.

3:42So check out ridehomefund.com for more info. But essentially, this was almost inspired by Chris. I've been running the Ride Home Fund for two years, and Chris was always a great source of deal flow. And my remit to him was always like, if there's somebody coming off the bench that had a previous exit or that you've known for a while that by definition you'd invest in whatever they were doing next, send them my way. and around December of last year, it all became AI stuff. And these were folks for whom this wasn't their first rodeo, who didn't bite on Web3 or crypto or whatever, but were activated by this AI moment.

4:22And we can give you examples if you're interested. But that also activated Chris, which he can speak to, which is we were both like, we want to back these people. The talent of our, you know, we've both been in the Valley for over 20 years. The talent that we're seeing coming off the bench activated by this moment, we just wanted a fund to back those folks. Okay. The real question is, where's the investment opportunity? I mean, that's, so of course you're on the phone with Mark Andreessen and he's like, yeah, of course there's opportunity there. But again, what everybody's saying about AI investment is you have the big companies and they're going to crowd out any seed company that you're funding that doesn't have.

5:02for instance, like the money to spend on the GPUs, or potentially they're building a thin layer on top of something like a chat GPT and chat GPT could just subsume it. For instance, character AI, right? Which is like, it allows you to chat with all these historical figures. Like it's very easy just to go into chat GPT and be like, all right, chat GPT or Bing or Bard, you're Abe Lincoln, let's have a conversation and it will do that. So then what's the value of character. So where are you guys seeing specifically outside of the people, the investment opportunity to actually make a splash here with seed companies and Series A companies that some of these other luminaries like a Keith or Boy or a Sam Lesson are saying do not exist?

5:44I think the skepticism or the reluctance from a lot of folks is exactly the moment to lean into this because now is when we're going to be learning what types of products actually make sense. So what Brian and I ultimately are looking to invest in is the productization of these new techniques that start from an assumption that software and technology, and that the way that humans actually use computers is going to go through a rapid revolution, but that those existing incumbents are going to defend their existing models and how they make money currently. Whereas a bunch of other folks who don't have to worry about that can build something entirely new that are set with new assumptions.

6:24And so part of that is looking at behavior. That's a big piece of how we learn to build social software that people actually wanted to use. And then the second part is coming up with generational brands that actually reconstitute existing marketplaces or product experiences that were in the marketplace, but that can be remade through technology. So looking back on things like Airbnb and Uber, suggest that these were ideas that people understand. You have property, you want to rent it out. You have a vehicle, you want to lease it out or you want to drive people around and charge people for that.

6:53But bringing technology like GPS and mobile phones to that allowed for these large companies to be created. And I think in a similar way, we're going to see the same thing with AI and generative AI. So I'm being reminded now, like hearing the way that you guys speak of this stuff, I'm being reminded of a Benedict Evans tweet where he talked about the answer is not that you're going to start to talk to Microsoft Excel. It's that the way that you do spreadsheets just changes because you into a completely different form factor where you're going to upload, for instance, the data. Exactly. Why do you need a spreadsheet?

7:22You'll just upload the data and talk to it. I would use the analogy of like, if this is a paradigm shift in compute, which we believe it is, what was the thing in the paradigm shift of the PC era that sold PCs was the spreadsheet, something that didn't exist before that paradigm shift that allowed accounting to 10, 100x or whatever. We're obviously looking for those things, but let me back up for a second. fundamentally what we believe is like you asked the question like well will the incumbents or the people with the large largest models always win if that ends up being true then yes this is academic but we don't believe fundamentally that's true we believe that fundamentally this is still software like how did zoom win over other sort of um you know um conferencing software why does this one win over that one, Figma win over that.

8:09Like in the end, it's the end result of what you do with the software for, um, for users in terms of their productivity and their delight in the product for all of the talk about generative stuff and AI stuff in the end, it's software. And in the end, it's inputs and outputs, right? So we have this theory that while everyone is still looking for, um, what the, you know, investing in this is not on rails yet in the same way that has been for things like SaaS or like D2C companies that people understand this is how you build a company that reaches this many users or customers or something. But in the end, we think that on the input side, it matters.

8:53And Chris will go into this about our theory of AI varietals, but it matters the models. It matters the data, the input that creates the things, and it matters the output. So you could have three different startups that are like, we use AI to look at x-rays and find cancer, right? But what have you trained that on? And does one model work better than the other? And already people are talking about things like accuracy and not having hallucinations and stuff like that. So people that get more accurate will win. But also, what have you trained it on? Do you like the results it gives you? Does the output, well, this output is more accurate, but that output in a clinical setting allows me to interact better with my patients and things like that.

9:36So in the end, we believe that this productization is the way to go because it will be the inputs and the outputs with the underlying software and AI stuff sort of abstracted away that will matter. And the analogy that I have to give Chris credit for coming up with is our theory of varietals like wine. Hold on. We're going to get to your theory, but I also want to ask, isn't this kind of a tough area to invest in because, you know, what is an AI company? Like if I'm investing in social media, or I'm investing in SaaS, okay, I know what I'm doing. If I'm investing in AI companies, all of a sudden, you're the range of companies like you guys can be consumer, you can be business, you can be anything.

10:20So how do you pick what you invest in? And then we can get into your varietals theory. It's back to people. It's back to application of the productization. Can I give you an example of a company that we actually have cut a check for. Wait, you guys have already cut a check and you haven't even finished raising your fund? It's a 506C. We're allowed to do that. We're moving fast. That's the whole idea. That's cool. Yeah, let's hear it. Yeah, definitely. F-I-Y.A-I, fixityourself.ai is training models on all of the world's sort of owner's manuals, user's manuals, specifications, and things like that.

10:59The idea being, Train an AI bot to fix your dishwasher, repair your car. My printer all of a sudden isn't working or is flashing this green light. What does it mean? As opposed to going in and finding the PDFs or calling the 1-800 number, they're training on every product under the sun. they're refining the model in terms of can this be a white label thing that a samsung or uh or a frigidaire or whatever would use to sort of be the the sort of bot that would enter interface with the customer so that the customer doesn't have to call the 1-800 number so that it would save on things like um you know uh warranty stuff or whatever or also on the consumer facing side what if there was one bot that anytime something broke you could use your smartphone to like turn on the camera and be like, okay, the dishwasher isn't working.

11:52Look in the back there. Oh, that hose is disconnected. And the bot tells you what that means and X, Y. So this is inputs and outputs. We love the founding team. They have previous experience in the space, relationships with OEMs and things like that. That is sort of a definitional example of the productization that that we like. It's not the only stuff we'll invest in, but like, that's, that's where we're seeing stuff that we get excited about. Yeah. I like that. I mean, the idea that I could like, for instance, you know, converse with my refrigerator to try to figure out like how it's broken.

12:28Yeah. The analogy I use is like, you know how every website has like the sort of chat bot on the bottom right hand corner that if you have a question like the AI, well, imagine a chat bot AI that comes into the real world, into meat space, not just on a website, right? Yeah. I think Elon's working on some of those. We just had Ethan Mollick on the show. He's a professor at Warren who assigned his students to use ChatGPT in class and talked about how lots of students are going to be cheating with this stuff. And I personally cannot wait for the first person to cheat with Chat Refrigerator and have the paper include the term.

13:07As an AI refrigerator, I cannot complete your paper, however. So let's hear about this AI varietals theory. So what does that mean? So to, I think, generalize from the example that Brian just gave, when we were going through this exact exercise, when we were trying to decide, you know, do we want to do this fund or not? And why now? And why us? It stood out to me that there's been this refrain, or refrain rather, in the business that data is the new oil. and it's been bandied around for quite a while. And the assumption is that if you have just a load of data, you've got some data lake or warehouse someplace on the internet or in someone's cloud, that you have an advantage for both training and building AI products.

13:54And while there may be some truth to that and having an enormous amount of data can be useful, there's also a lot of crap out there. There's a lot of stuff that's undifferentiated. That's just kind of nonsense. And having worked at various tech companies, I know for a fact that a lot of the stuff inside of these spreadsheets or whatever just is not structured in a way that can be useful for training these data sets. So as I was thinking about a better way to think about the opportunity, it occurred to me that it's not that data is the new oil. It's that data can be thought of as a type of terroir.

14:27And, you know, obviously coming from California, being in California now, you know, we've got Napa Valley right up to the north. And so there are all these places throughout California where the terroir, the earth, the place in which grapes grow, represent different types of cultivation opportunities. So essentially by cultivating the land and growing grapes in it and then harvesting those grapes, which, of course, have grown in different climates with different soil types, that produces different types of expressions in wine. And those different expressions in wine, if you bear with me for this metaphor, are kind of like the types of products that we expect to be able to create from creators of these products that understand deeply the data that they're working with.

15:09Maybe they've actually set up the data ingestion pipelines and the refinement pipelines. They understand deeply a customer segment and the needs of those customers. And perhaps most importantly, they have the vernacular or the language of that customer, and they know how to speak that customer and to deliver products and services that fit into the environment or the industry that their customers are in. So it's not enough to just have a whole shit ton of data and just to build a company or product. You actually have to have some empathy for the market that you're dealing with. And we think that that's going to be the opportunity for founders that are thinking about that and are thinking about the whole life cycle of building AI-powered, AI-enriched products, that there are going to be opportunities.

15:53And these opportunities are going to take many, many years to play out. Well, many, many years. In my thinking, it's like three to five years, maybe eight years. But nonetheless, now is the time to start. Now is the time to learn. now is the time to experiment. And for those founders who either have left big tech companies or more importantly, were laid off or fired from those companies, they're going to be hungry to be building those products and those services. So that's really where I think Brian and I, both students of history, both internet history, as well as social media history, as well as longer term history, are seeing opportunities for us to support those founders, coach them through launch and help them get distribution and to be discovered.

16:38I also think, Alex, that it is going to be there's going to be a ton of M &A over the next three years. Right. So I, like you, have seen like the studies and things like that. And anecdotally, oh, we didn't expect that all of the incumbents would adopt this AI stuff that quickly. Right. Like everybody and their mother has added AI features to their existing products. Right. Where it gets tricky is where it gets fundamentally disruptive to like, again, Microsoft adds AI to Excel. But what if you had something that I could do that obviates the need for a spreadsheet entirely? Right. Right. The thing that we're looking at is like, again, to use the spreadsheet example, like Chris and I believe me, there are a hundred different startups out there that are like, draw on a napkin a room and the AI will generate like sort of the perfect interior design.

17:43A beautiful interior for you. Yeah. But like my wife is is an architect. And every time those come through, we float them by her. And she's like, can it create a BIM file for me that I can send to the structural engineer? No. Why? Because it's two kids, 25 years old from Stanford. There's no architects on the thing. You know, she's like, you know that we have I'll lose my license if things do not meet certain like building code standards and things like that. So like our like our dream sort of company to come our way right now would be like a 25 year old from Stanford. That's an AI guru, an architect with 20 years experience that understands, you know, the needs of the profession and the and someone that's ex Autodesk.

18:30right? Because what my wife is like, is like, the day that you could come to me, and you could train my you could train an AI on my previous, you know, career long history of drawings. And like, this is how I like to design an interior. This is how I like to, and, and, and you could do the napkin sketch into the AI thing. But it's based on what I like to do. So it's like my style of design and it's functional, like so that in a day I could have BIM files that are actionable. That's the 10x sort of equivalent to the spreadsheet. I think maybe, maybe Brian, another way to layer this is, is I do like Microsoft's frame of co-pilots.

19:11And if we think about these as tools for collaboration and for reasoning and for thinking through or going through the creative process, then ultimately you kind of want to mix where, you know as as brian said like lisa would submit her designs to train an ai she would produce a series of sketches she would then submit them to the ai to get some other ideas or other directions perhaps returning sketches in her style in order for her to work out the thought process and then when it's time to actually move forward into the finalization workflow then they could she could go into a mode that is all about that type of collaboration the problem is that so many of these generative AI tools are designed only on the final product and the outcome.

19:53And those outputs and outcomes are based on a popularity contest on the internet, which is all good and well, but not necessarily the most appropriate for coming up with novel ideas or synthesizing new outcomes or coming up with entirely new concepts or practices. So it's great for maybe returning to things that have been done or are well understood in the marketplace and are therefore or somewhat commoditized. But creating things that are new expressions and creative, I think, still is an area where there's loads of opportunity. Yeah, this is interesting for two reasons. The first is that it seems to be an application that something like a Google will not want to go after because it's just small enough to be under the radar for them, but just big enough that could be a really interesting company.

20:38And the second reason is that because we talk about generative AI, and it seems like the conversation always goes toward chatbots. But if it can go towards, for instance, blueprints or having a dialogue with other like important technical but creative images, then you could really, and plans, then you could really get into an interesting place. We're here with Brian McCullough and Chris Messina. We're running this both on the Big Technology podcast feed and on the Tech Meme Ride Home feed. If you're a Big Technology podcast listener, I definitely think that their show is worth checking out. Brian's show is worth checking out every day.

21:11A little update on what's going on in tech. And if you're a Tech Meme Right Home listener, I'd say, you know, come give Big Technology Podcast a shot. We do weekly, twice weekly interviews. Oh, I'm sorry. I was going to say, how many times have we seen on Twitter people say that Tech Meme every day and your show for the in-depth sort of stuff twice a week? I think it's a great pairing. It's been, it's really, it does, if you're going to think about your tech podcast media diet, you know, I think putting those two together, that's the way to go. All right, let's take a quick break. We'll be back on the other side of this break to talk about how these two were able to get in front of the Mark Andreessen's and Dennis Crowley's of the world.

21:51And we're able to convince them to invest. And then we have a couple of AI related stories that we want to talk about. AI warfare, AI in the Hollywood strike. We'll make it a quick second half. So stick with us. We'll be back right after this. These days, it feels like every dollar should be working a little harder. But figuring out where to put your cash can be confusing. That's where Wealthfront comes in. Wealthfront is a tech-driven financial platform built to help you grow your savings into long-term wealth. Their high-yield cash account through program banks offers a 3.5 % APY on your uninvested cash as of November 7, 2025.

22:25And there are no monthly fees, no minimum or maximum balance to earn that rate, and you can even make free instant withdrawals to eligible accounts in just minutes, 24-7. So your money can always be within reach. Right now, Wealthfront is offering new clients an extra 0.65 % APY over the base rate for three months on up to a$150 ,000 balance. That's a total of 4.15 % variable APY when you open your first cash account. Go to Wealthfront.com slash Big Tech to sign up today. This is a paid testimonial for Wealthfront. It may not reflect the experience of others, and there's no guarantee of future performance or success.

22:58Wealthfront brokerage is not a bank. Rate is subject to change. Promo terms and conditions apply. For more information, see the episode description. Shape the future of enterprise AI with Agency. A-G-N-T-C-Y. Now an open source Linux Foundation project, Agency is leading the way in establishing trusted identity and access management for the Internet of Agents, a collaboration layer that ensures AI agents can securely discover, connect, and work across any framework. With Agency, your organization gains open, standardized tools and seamless integration, including robust identity management to be able to identify, authenticate, and interact across any platform.

23:41Empowering you to deploy multi-agent systems with confidence, join industry leaders like Cisco, Dell Technologies, Google Cloud, Oracle, Red Hat, and 75-plus supporting companies to set the standard for secure, scalable AI infrastructure. Is your enterprise ready for the future of agentic AI? Visit agency.org to explore use cases now. That's A-G-N-T-C-Y dot O-R-G. And we're back here with Brian McCullough and Chris Messina, two investors who are raising a$15 million target fund from investors, including Mark Andreessen and Dennis Crowley, Chris Dixon. I have some thoughts about Mark Andreessen that maybe I'll share on a different show.

24:24But in the meantime, I'd like to hear how you were able to get in front of him to raise the money from him. And doesn't he also have like an AI related fund? So talk a little bit about that process. Yeah, sure. The short answer is that it's because Mark and Chris Dixon listen to the show. Chris is very kind. He has said multiple times that when the Andreessen Crypto Fund invests in a company or hires new people, they give them my book. so the bottom line is is that actually it was it was Chris and Mark those were the first two LPs we reached out to there's some other luminary names that you would know but have not agreed to allow us to use their names to fundraise essentially we went out and and we said sort of what you said at the beginning which is no one knows what this space is come in with us and you As an LP, you'll be able to see the people we're seeing and the companies and ideas that we're seeing.

25:25Great. So we have a few minutes left together. I thought that we're doing this as a special episode, but we still want to cover some of the news. I feel like doing a big technology podcast or a Tech Me Ride Home show without going over some headlines would be a complete waste. So let's not waste this opportunity. Now, we're in the middle of this big writer and actor strike inside Hollywood. and we put a document together to talk about, you know, to sort of discuss the type of stories we wanted to cover. And, you know, usually I like to go through like a more robust setup, talking about the story and what's happening and then ask a pointed question.

25:59But there's a one-liner in this document that just said, AI and the Hollywood strike, Brian's anecdote, without any detail. And I was like, I'll be damned if I don't ask about that. So, Brian, let's hear your perspective on what's going on here. And it seems like you've had an interesting interaction. so I you know I do daily tech news and do headlines like you do headlines on your show so you might have done a similar piece I think it was in the Washington Post maybe where they were talking about how um in this sort of strike moment where how long has it been going on more than two months now um like this might be the moment where TikTok stars the writers become the writers and the now the actors.

26:41Yeah, that's shorter. Yeah, go ahead. So, but that studios might gravitate towards TikTok stars since they're not in a union to in the same way that in the strike in the aughts, that sort of begat the takeover of reality TV because it was cheaper and non-union. And so TikTok stars might be moving over to fill the gaps of what Hollywood needs for content. At the same time, the article posited that this was training folks that were in Hollywood, that were actors, writers, or whatever, to be like, I need my own platform. It almost pushed them into the creator side of the equation. And so the article was positing that this could kill both sides of the equation in terms of the traditional Hollywood model.

27:28And so my anecdote is this. You know, I'm in podcasting. I talk to a lot of podcasters. I also weirdly enough have a background in comedy and stand-up comedy and improv and stuff like that so a lot of the shows and friends that I listen to their podcasts are like comedy podcasts but let's just say a broader picture people that work in Hollywood right all of these years they've had podcasts like comedy and various Hollywood folks were early adopters to podcasting because it was like a supplemental thing. Like if you can get staffed on a writer's room for a couple years, then you make a decent living, but then you might not be staffed on a writer's room for a couple years.