We're In An AI Bubble. Here's The Upside Nobody's Selling. | Rachel Chalmers, @Generationship

If you're a female founder, you should find some male founders to join your team. Otherwise, your chances of getting funding are near zero.

Honestly, it's advice that I've given because only 2% of venture capital goes to women. Statistically, your chances of raising capital do increase if you have a male co-founder. Probably 3 quarters of the companies in our portfolio have male co-founders. But I will say that the most successful ones are all women teams. We are not a charity. We are a financial fund. We have now 15 years of documentation that women-led organizations actually outperform male-led organizations by about 43%. There's a huge disconnect between how the funding is being allocated and how founders are actually performing. That disconnect is our opportunity. We invest in overlooked technical women founders, they outperform the market, and we make money. It is not an oracle, it's not intelligent, it's not conscious. In any meaningful sense. All of this delusion that people built around it, that you could cut the government in half and fire all of your workforce and get rid of all of your institutional knowledge and AI would make up the gap, is already becoming obviously a very bad bet. My absolute best tip when you're fundraising for anything is just go out and collect 100 rejections. Your job is to collect rejections. If they tell you, no, that won't work, that's a positive. piece of data. Collect those rejections. Map out the negative space. What I tell my kids is what I tell my founders. Find that weirdness. Find the thing that is completely unique about you. Because all of the grinding stuff, LLMs can do that now. Work for humans is caring about stuff. It's making intuitive lateral leaps. It's taking care of things, having novel thoughts. Almost by definition, LLMs can't have novel thoughts.

Today I have Rachel Chalmers, she started her new VC fund, Generationship, with Michelle Yee, and just recently closed their inaugural fund last month, oversubscribed at $2.7 million. So today she can teach us how to be better founders with her over 12 years of judgment and observation as an investor. So welcome, Rachel. Thank you.

Thank you very much. I will do my best.

Awesome. Do you have a version of how you introduce yourself that's, you know, AI cannot find online or not on LinkedIn?

The way I usually introduce myself is that I'm a failed English professor. I was at graduate school in Ireland, actually during the 1994 World Cup, and I was planning to be an English professor, and I went over to some friend's house to watch the game, and to this day, I'm the only Australian who can't even pretend to be interested in sport. So, I got bored, even though Ireland won that game, and I picked up a magazine off the floor, and it was lime green and orange, and it was the first time I'd seen a copy of Wired magazine, and I started leafing through it. And by the time the whistle blew to end the match, I was no longer going to be an English professor. I was going to go to San Francisco and write about computers.

What an amazing story. Yeah. It's hard to believe, right? Life is changing. You know, in an unexpected moment.

Yeah.

Yeah. And you're just gonna go with it.

Yeah. Even then I could see that tenure in English departments was going to be increasingly difficult to get. So, you know, the writing was already on the wall for what we've seen play out over the following 30 years with adjunct professors and people, friends of mine cobbling together careers. With 2 or 3 contracts at 2 or 3 different institutions, and it's no life. And I'm a horse girl, so I wanted to have enough money to, to keep a horse. And that meant finding something more lucrative than teaching English.

Do you still have a horse?

I do. His name is Lenny. I rode him this morning. He is a Mustang born wild in the desert of California. And these days he makes his living as a fancy little dressage pony.

Oh my goodness. I feel like I need to ask you and then just have on a side, side podcast about horse riding. I took a horse riding class and I, I, can you tell me that you can tell someone's personality how they can ride a horse?

Oh, the challenge is not getting me to talk about horses. The challenge is getting me to stop. But one of the fascinating and lovable things about horses is that a lot of what you think you know about horses is wrong. Like we have this spirit stallion of the Cimarron stereotype of like the brave stallion leads the herd and the mares just follow along. That's the opposite of how it actually works in the wild. I've been lucky enough to see wild horses' behavior. What they actually are is a gestalt entity that's led by whichever lead mare is paying attention at the time. So a little bit like meerkats. One of the horses, its head popped up and it's scanning for predators. And then when they do see a threat, I was watching the Montgomery Pass herd, which is actually preyed on by mountain lions. It's the only herd in the United States that's managed by mountain lions. When they do see a threat, they move like a school of fish. Like the lead mare startles and all of the other horses have one ear on the lead mare. And so they all move together like starlings, a murmuration of sparrows, and they move as one entity. What's really interesting about that is they have all of this body language, which is about connection and mirroring and Empathy, and they're perfectly happy to incorporate humans into that gestalt. So if you're doing it right, if you approach a horse with the right combination of confidence and humility, you become part of its herd and you get to communicate through that same body language that horses use to communicate with each other. And for humans, particularly nerdy humans who tend to be brains in jars locked up in our own heads, This opportunity to form a relationship with another creature and become one half of a whole, it's pretty magical.

Do you test your potential prospect founders on riding a horse before you—

No, I would never do that. It would be unfair to the horse.

All right.

You know, I even dragged my children along for riding lessons and neither of them are horsey. So, It's such a challenging and dangerous sport. I've broken both ankles and my collarbone at different times that I would never force anyone else to do it. But it is one of the great joys of my life.

Great to learn about your experience of riding horses. I'll definitely ask you for coffee for more chats around horses. Talking about founders, I reached out to Michelle about a month ago. Because I saw her beautifully written LinkedIn post. She has firm belief in that post in women and women plus technical founders. I've been talking and interviewing some other investors who sit on the complete opposite side. And I remember one of them said something like, if you're a female founder, you should find some male founders to join your team. Otherwise, your chance of getting funding are near zero. You must have heard something like that too, I'm sure.

Oh, won't you stop?

Constantly, constantly.

And honestly, it's advice that I've given because only 2% of venture capital goes to women. Statistically, your chances of raising capital do increase if you have a male co-founder. I do think it's incredibly important to have a co-founder. I think getting somebody else to quit their job and go into business with you is the first test of anyone starting anything. Probably 3 quarters of the companies in our portfolio have male co-founders, but I will say that the most successful ones are all women teams. Yeah, there's this bifurcation, to greatly oversimplify this, there's an increasing bifurcation in venture capital. You have a bunch of funds that have historically been incredibly successful and are now raising many billions of dollars and have billions of dollars in assets under management. And they're increasingly doing consensus investing. They're investing in the top companies that come out of Y Combinator. They're investing in repeat founders, like the founder of WeWork. basically writing blank checks to people who've done it once and will do it again. They're behaving less and less like venture capitalists. They're not taking any significant risk. They're doing deals that are obviously good and they're marking up their own deals. They're kind of in a different business to what we're doing at Generation Ship. At Generation Ship, we're pursuing what was once the classic venture model, which is looking for opportunities in the market where demand for a product will rapidly outgrow the ability of the business to bootstrap itself. So, we're still looking for that hockey stick growth. We're just looking for it outside Stanford and outside Y Combinator. That said, Y Combinator is a great program where we do have a Y Combinator company. We don't believe that all of the merit in Silicon Valley Belongs to white men. I mean, we have now 15 years of documentation from organizations like McKinsey and First Round Capital and Female Founders Fund that women-led organizations actually outperform male-led organizations by about 43%. There's a huge disconnect between how the funding is being allocated and how founders are actually performing. That Disconnect is our opportunity. We, we find a whole bunch of incredibly overqualified, brilliant founders who are being overlooked by the market. And with just a little bit of capital and encouragement, they outperform. And, and that is the business that we're in. It's traditional venture.

You said you made one critical decision early on to invest in these women founders, women plus founders. And you also mentioned that they are a population who are chronically underpriced.

I find that very interesting.

It's a totally different take from a diversity fund, right? That's not how you market yourself.

We are not a charity. We are a financial fund.

Can you walk me through this actual mispricing? Like what does the market, what's the market getting wrong?

I think in the vast majority of cases, there's not even any malice in it. There is a huge tendency on the part of investors to want to invest in younger versions of ourselves. And for all of the men named Steve and Cameron, that means investing in younger men named Steve and Cameron, to again, grossly oversimplify. Michelle and I, in very different ways, have had absolutely blessed trajectories. I was lucky enough to be one of the founding employees at 451 Research. I covered over 1,000 venture-backed companies, and that was my basic training for being an investor, was just pattern recognition at large scale. Michelle, I don't know how much you know about Michelle, my amazing amazing partner, but she grew up very, very poor in Seoul, in the shadow of Gangnam, which is the very wealthy region. And so, she would walk across to the Gangnam Public Library and she taught herself English out of library books well enough that she got a full-ride scholarship to a US college when she was 13 years old. And by the time she was 16, she was—

That's crazy.

Yeah, no, it's ridiculous. By the time she was 16, she was working at IBM and she worked her way up to become the lead engineer on Watson, IBM's AI platform. So, we're both like way out on the very narrow end of the normal curve in terms of luck and circumstance. And that gives us a very different set of patterns to match against from all of the guys who did a Y Combinator startup and turned around and invested in other guys doing Y Combinator startups. We came from really anomalous backgrounds, and that lets us see potential in other founders that come from really anomalous backgrounds. And I do think that's where the real returns come from. These incredible founders are getting systemically overlooked by the traditional venture firms, which means we buy early, we get a lot of ownership, and once they're able to prove their potential in metrics, in ARR, in, you know, the actual fundamentals of running a business, they get these huge markups. So, we already have 2 markups in the portfolio. We've only been deploying capital for 2 years. We wrote a very early check into Adira. We introduced those folks to Kevin Crosby at GitHub, and GitHub— Kevin took them upstairs to M12, Microsoft's strategic venture arm, and we got a 4x markup on Adira within 3 months. We also invested in Coval. That's the Y Combinator company I mentioned, led by Brooke Hopkins, who's incredible. She ran the test harness at Waymo testing the autonomous driving model, and she realized that Exactly the same sort of large-scale statistical evaluation would be needed for voice agents. And so, she went to Y Combinator to start that company. She's just also seen an enormous markup from where we invested. So, but by the time these women get the opportunity to prove themselves in the market, we see that 43% outperform and you get those markups at Series A and Series B. We just get in early when it's not obvious.

Does this underpricing meaning you can, you know, you can invest in this group at a discounted valuation? Am I understanding it correctly?

Yes, effectively. So the way venture works in an ideal world is venture is a series of gates and each gate corresponding to a stage or a series represents the de-risking of the proposal. So we invest super early. We like to be the first check-in. Which means that the risk is sky high. But we think investing in women lowers that risk more than other investors think it does. By the time you get to seed or Series A, you're supposed to have product-market fit, maybe some repeatable sales. By the time you get to Series B, you're expected to have, you know, a machine that systemically reduces customer acquisition costs and increases long-term customer value. And each of those gates is designed to represent a reduction in risk. And at each of those gates, you're supposed to bring in a new lead investor to reprice that company against the market. So where we get our big markup is we jump in early because we think the risk is not as big as all of the other investors think it is because we think women are brilliant and competent. We get to seed and Series A and the other investors finally have some metrics to look at and chew on and they're like, This is actually a really legitimate company. And so, they come in and they reprice the company. So, if we bought in at a $10 million valuation and Microsoft comes in and says, this is a $50 million company, that's our 5x markup.

Can you share a story of any of your recent investments that you think the woman founder is an anomaly and what made you write that check? What did they say to you in the meeting that made you decide, okay, We're going to invest in this team.

So, I mean, one of the ways in which my career was particularly blessed is that I've actually been a very early investor in 3 women-led companies. LaunchDarkly, which was founded by the amazing Edith Harbour, who runs the Western States Trail, which is 100 miles up and over the Sierras. Aviatrix, which was founded by the astonishing Sherry Wei. She's a network engineer of considerable genius. We were actually looking for a name for the company, and I came up with the name Aviatrix in honor of her. And Honeycomb, which is a couple of Facebook veterans, Christine Yan and Charity Majors, who is a dear friend now. They were all just incredibly technical, accomplished women with a very specific vision for how the future could be better. Edith wanted blameless postmortems to be easier to do. She wanted incremental changes to large systems not to bring down the whole system. if something went wrong. So she invented feature management so you could toggle flags off and on. Sherry wanted networking to work sanely. She wanted the networking software to be well-designed and give a good user experience. Charity had worked with an internal tool in Facebook called Scuba, which is a huge columnar data store that let you ask open-ended questions of how the infrastructure was behaving so that instead of having to set thresholds and metrics and alarms to let you know when the infrastructure was behaving out of bounds. You could just say, well, what is happening in the infrastructure? Rather than being reactive, you could be proactive. So, I had those patterns to match against. And when I met Emily Long and the other founders of Adera, Ariadne and Alex, I already knew that Kubernetes and large language models were going to need a new version of virtualization. I had seen XenSource and VMware and Docker come into the market and reinvent how large systems were deployed. And so, I was looking around for something that would do the same for Kubernetes and large language models. I met this team and they were like, the way to do this is with the old Xen paravirtualization method brought up to date in the Rust programming language. And I'm just like, yes, that fills a gap that I already sort of instinctively felt was forming in the marketplace in a way that's technically incredibly powerful and mathematically simple. And so, we invested in Adera. Microsoft marked it up. The company's doing great. That's not even the one I want to tell you about. A few months ago, Emily pinged me and said, my friend Gabriella is looking to do a startup. She's spinning it out of Informal Systems. Can you talk to her? Got on the phone with Gabriella, and what she's doing is formal verification. Which has always been mathematically possible with software. You can construct mathematical proofs if you have a well-enough written spec. You can definitively say whether or not the software matches that spec or not, whether the behavior of the software does what the author had intended. The weasel word there, of course, is given a sufficiently well-written spec. Writing that spec has historically been a gigantic pain in the butt. What if you had a magic machine where you could tell it what you meant and it would generate a spec for you? LLMs suddenly make it much easier to write those specs. And therefore, it's a huge barrier to entry to perform more verification. So, Gabriella explains all this to me. You know, this is a niche market that could actually suddenly explode and become a really essential step in the software lifecycle where you have 90% of developers now working with AI to generate code, and only 40% of them trust the output. What's missing? A formal verification step where you can say, does this actually do what I meant it to do? So, I talked to Gabriella. I got Michele to talk to Gabriella. We were both excited, but neither of us have a strong background in formal verification. Luckily for me, I'm married to a guy who does. He's the Rust guy at Meta. And so, I got Jeremy on the phone with Gabriella, and I was actually on a different call when he got off. And my husband is a very, he's very dry and he's very straight-faced. He has a great sense of humor, but it's very low-key. But he gets off the phone and I'm on this other call and he's standing in front of me with this big grin and 2 thumbs up. And I'm like, yes, we found another one. So, we've invested in Gabriella's company. It's the most recent investment that we've done is Generation Shift. It's called Quint. It's on the web at quint.sh. Formal verification is going to be the new, new thing. You heard it here first. We're looking at another company in a related sector that we're also very excited about. So yeah, Gabriella and Quint are the new ones that I'm incredibly excited about.

I want to understand what are the traits that you're seeing in these winning teams that made you bet on them? Is it just the market and the use cases?

What about the people? There are certainly traits that I look for. You know, I mentioned Edith and Sherry and Christine and Charity. They're sort of my model for, these are the kinds of technical women who build unicorns. Some traits they all share and that they share with Brooke and Emily from Koval and Adira, they're all able to articulate this contrarian view of how the world can be. I guess I'm a completely unreconstructed '90s tech idealist, like, Back in the days when Wired magazine was orange and lime green, we all thought that the internet was going to be a great thing and it was going to empower a lot of people and it was going to change the world for the better. My views on that have been significantly tempered over the years, but I still do think that the point of having a great big human brain and monkey hands is to try and improve the world a little bit, to try and leave the world a little bit better. than you found it. And I no longer believe in gigantic incremental revolution, gigantic step change revolutions, but I do think incremental improvements to things add up over time and make things a lot better for ordinary people. And despite of all, all of the bad news, you know, Mm-hmm. We've managed to eliminate smallpox. We've greatly decreased the amount of malaria and the amount of tuberculosis in the world. We're We've made huge strides on Guinea worms. In my own life, my blessed career is enormously more privileged and interesting and engaging than my mother's career was. And I hope the same for my kids. And so that kind of like, not the grand revolutionary stuff that the big firms are trying to sell you, but the day-to-day incrementalism of let's find a problem and let's solve it elegantly. That's really important to me. So, when I find technical women who have that contrarian view, who can argue it really passionately and really well, who can become the leaders of a movement to try and do things a bit better, let's create observability so we understand our systems a bit better, let's do software networking so that our systems interact with each other a little bit better, those kinds of leaders are the nodes around which positive changes can form. And one of the reasons I love investing at the infrastructure layer is because it's so highly leveraged. One of the knock-on effects of Ida's success with LaunchDarkly is that if you have a team that's using software that supports blameless postmortems, then nobody's pointing fingers. The team is a lot less political. You're actually working together to find the truth and to make the situation better. And you're not just fighting. And wasting energy in interpersonal conflict. And so that's sort of my very quotidian day-to-day big vision of how the world can be a better place is let's find some problems that the new layers of technology let us fix in more interesting ways. And let's fix those problems. And then let's move on to the next set of interesting problems that the new technology we just built lets us solve in more interesting ways. Software is archaeology. We build cities on top of cities on top of cities, and the lessons that we learned in old cities come back into the new cities, and gradually everything just gets a little bit better over time.

Do you think we as women have any advantages over men founders? Oof.

Who's going to listen to this podcast? Yes, I absolutely do. In my career, I ended up writing about infrastructure software because it's hard and it's boring, and it's not flashy and it's not glamorous. And I look around at all of my women friends who survived in the industry, and we all made similar choices. Charity made a similar choice. My friend Diane works at a hospital up in Portland. We went for the boring jobs so that it would be harder for people to fire us. Remember, I was an English professor. I wanted tenure. I moved into the technology industry. I wanted to be hard to hire, hard to get rid of. I wanted to know stuff that would make it difficult for people to lay me off. I have been laid off twice, but I'm still here. And women make those choices. Women choose less exciting stuff because they want to have a more durable career. We, our reproductive lives are very different from men's reproductive lives. Our Risk-benefit analysis are very different. That's changing all the time. But yeah, I think we have experiences that make us more resilient, that make us more aware of a multiplicity of voices around the table. We have fewer blind spots because we have experiences that are different from men's experiences. And these are all generalizations, but they've borne themselves out. And I will say that women of color have even more advantages than white women. in the calculus that I'm doing. And men of color have more advantages than white women.

Why?

I, as a white woman, am only the second most privileged set of identities in the world. But the more intersectional challenges a person lives with and learns to negotiate, the more double and triple and quadruple consciousness they have, the more subtle and nuanced their world model is. And I think that's an enormous and underrated advantage that women and people of color and queer folks and neurodiverse folks have as founders.

There's an interesting collision between part of us, we want like stability, and part of us have actually experienced a lot of diverse experience throughout our career. and what we're pursuing, right? I came from Tibet, so grew up in a very impoverished school district. I can relate to Michelle on what she experienced.

And my mother was working class. She was from the north of England. She was the first in her family in, you know, 1,000 years, the first in her family ever to go to college. And she had a nervous breakdown and washed out of college. But I think, you know, when I go to industry conferences and I'm sitting there in the room with Jill Biden or Gwyneth Paltrow, It's my mother looking out through my eyes saying, this is all tinsel. It's not real. The real work is that incremental problem solving. It's finding somebody's challenge and making that a little bit easier. That's the real thing. And I think that's easier for us to stay grounded if we don't come from inside the Silicon Valley bubble.

But we want both, right? We want this incremental mentality of trying to improve, make our lives better. bit by bit, but also we are adventurous, we are explorers. You kind of have to have both worlds to become a successful founder.

Yeah, it's a paradox.

It's weird, isn't it?

Yeah.

And do men want more of a bigger leap than women?

I don't really know much about men. I mean, I know a lot about men because I'm a woman living in a man's world, and I've read way too many books and watched way too many TV shows that were written by men for men. But honestly, for the last 20 years, I've thought way more about women than I have about men. One of my criteria for investing is interestingness, which is something about novelty and something about surprise and delight. And I think a lot of the narratives around Silicon Valley right now, and you know the ones I mean, are quite boring. So I don't spend a lot of time thinking about them. There's a lot of business books and biographies that I don't read. Um, because of my constant quest for interestingness.

So, can you teach me how to get mentally ready for fundraising as a woman founder in general then? Let's say if I wanted to build a product or, or if I'm already building, how do I know that?

My absolute best tip when you're fundraising for anything is just go out and collect 100 rejections. You know, check them off. Your job is to collect rejections because you will do call after call after call from, if you're going to be a founder, if you're going to be fundraising from, from angels and venture capitalists, all of them, co-founders, potential customers, just go and collect your rejections. So, customer discovery is the absolute crucible of doing any startup. If you think you have a good startup idea, go and do 100 interviews with potential customers. And if they tell you, no, that won't work, that's a positive piece of data. That's an important piece of data. That means that you're not going to go down a dead end. Collect those rejections, map out the negative space. And when you eliminate all of the things that are easily dissuadable for you, you end up with the core thing that you know to be true, that you're absolutely convinced of, that you can easily spend the next 10 years being completely obsessed about. And we went through this process when we started the fund. We were originally going to be this consulting firm accelerator. We were going to do all underrepresented founders. And we had a fantastic mentor who was just like, no, no, Narrow it down, narrow it down. What's the core of it? What's the core thing? What's the central thing? And we went out and fundraised for it. And the more we pared it down and the more we pared it down, it was like technical women are underpriced at pre-seed. And we're so absolutely convinced of that. We have a huge stack of research and it proves itself every day in our day-to-day life. That was the idea that was so good that we couldn't walk away from it. And to get there, you have to cut away all of the other stuff that isn't that idea. And that's what you do when you collect all of those rejections.

It's so interesting because you kind of have to map out the area, the surface, the map. And in the meantime, you kind of have to have passion, right, about what you end up getting. But what if like my passion lies in that area and I end up in something that has positivity, but what if I'm not interested in it?

People will tell you no about your core idea. Plenty of people have told me, no, women aren't any better at infrastructure companies than anyone else. And I just don't believe them.

You will get— How do you fight that? You are fighting it all the time, right? I bet you still have rejections today.

Oh, of course. Yeah. And I'm going to go and raise a bigger second fund and I will get plenty more rejections. I have so many more rejections in my future. I think the whole process is a process of whittling down to your core convictions. I think you find it by finding it. You find it by talking to enough people and looking at enough data until You convince yourself that you know something to be true, and that is your competitive advantage. Like, we're all in this market now where we're trying to get attention and we're trying to build a voice and build an identity and establish something that we're known for. First, you have to find out what you know for yourself, what you've learned over the course of your career, what you genuinely believe to be true and good and exciting. You've gotta find your own interestingness.

And to achieve that, I have to get 100 rejections.

Yeah.

And in the meantime, constantly reflecting on what I'm interested in. I think it's the hardest part about building anything.

Yeah. And it's also the most worthwhile.

So how did you go to market? How, you know, how did you land your first few LPs?

We got a few checks just from people who really loved me and wanted me not to get too discouraged at the beginning. And then we got a lot of rejections while I was still trying to land the right thesis. And then it was, I will always remember it because I was sitting out in the garden with my laptop on my lap and I was really scared about this call because I've known Marc Templeton, who's been the CEO of Citrix and DigitalOcean for a long time, and he's been a mentor and, you know, he is a good friend, but I also knew that he wasn't going to invest unless he thought it was a real thing. And we get on the call and I pitch him and he's like, yeah, I'm in. And I thought, this is actually going to work. Like, he's my first big name LP who's no pushover and is a really gifted investor in his own right. And I have got to the point where my story is good enough that he can see what I see. That was really the turning point for me. And I got a lot of rejections after that, but I got a lot of big name investors after that as well. And we closed oversubscribed and ahead of schedule. And I'm confident we will be able to raise fund 2 as well.

Why do you think he invested in you guys? What did he see during that call that you were scared of?

I, and this is the common thread across all of the LPs who said yes, I was able to paint a picture for him of a better world and a way to get there that he could believe based on his own experience.

Can you share with our audience what did that better world look like?

We invest in overlooked technical women founders. They outperform the market. And we make money.

Sounds easy and compelling.

I'm not a very smart woman.

Do you think we're in an AI bubble right now though?

Yeah, absolutely. AI, like every platform shift that I've lived in before, the web, the cloud, virtualization, Bitcoin, has a core of genuine interestingness. Bitcoin the least, it's like a big, slow distributed database, but the others have a core of like genuinely interesting and novel technology, and then this gigantic hype bubble around it. What has happened with large language models is a super, super interesting core that really lends itself to a lot of use cases, like upgrading legacy code and doing all kinds of transformations, like the formal verification stuff that was really challenging to do before. However, it is not an oracle. It's not intelligent. It's not conscious in any meaningful sense. And all of this delusion that people built around it, that you could cut the government in half and fire all of your workforce and get rid of all of your institutional knowledge and AI would make up the gap, is already becoming obviously a very bad bet. OpenAI in particular has invested incredibly heavily in infrastructure, which costs real dollars. And I was just reading an article before I joined this call about how Their projections that they'll make $100 billion in ads and they're struggling to make even $1 billion. And an expert thinks the total market for ads in chats, in chatbots is probably about $5 or $6 billion. So the numbers just don't math. However, what's really interesting is that I think all of this over-provisioning of infrastructure that we're doing gives us a legacy and it's not exactly like dark fiber, I know, but it's It's dark fiber with compute attached. Like, there's fiber in these data centers and there's compute, which will eventually go out of date. But suddenly the thing that has always made scientific computing hard, you know, that cycles are expensive and scarce, has gone away. You know, another barrier to entry has fallen. I think there's gonna be amazing stuff happening in high-performance computing and scientific computing over the next few years, mopping up all of this over-provisioning of That's my guess.

There's a part of it is very exciting and it's progressing. And there's the other part that people are thinking, when is it going to burst? What's going to happen? Are you concerned at all?

Of course. No, I live in the real world. I'm terrified all the time. I'm a sentient human. I have children.

In what ways do you foresee that it's going to pop?

So the thing that worries me considerably more than the AI bubble bursting, because I lived through bubbles bursting before. The thing that worries me a lot more is climate change. I just got back from London where it was 34 degrees and 34 degrees in London is absurd. It's completely ridiculous. None of the housing has air conditioning. People were dying. That's right. It was much, much hotter in France. You know, these kinds of temperature extremes, the wildfires in Canada, these were unthinkable. When I was a child, and we've dumped enormously more carbon dioxide into the atmosphere since I was a child. That's what really worries me. And what I see happening geopolitically is a bunch of petrostates and very wealthy people who want to maintain the status quo because fossil fuels are very dense and they're very easy to hoard. And the rest of us who would really like to transition already and who are excited about the dramatically falling cost of solar and wind and who don't care about the fact that those things are hard to hoard and are by nature Distributed and will cause, you know, much less centralization in the economy. And that dichotomy between a sort of authoritarian centralized view of the world and intelligence can come from anywhere. It can come from the slums of Seoul. It can come from Tibet. It can come from suburban Sydney, New South Wales, runs through all of the work that I do. It's like, do you want a future in which 8 guys have all of the money and the rest of us are used for food? Or do you want a future in which everybody gets access to a public library? Everybody gets access to compute. Everybody gets to live somewhere that's not too hot and not too cold. Everybody gets clean water. Everybody gets healthcare and education. I know which future I want. I know which future I'm working towards.

The second one.

Yes.

Yes.

It's the second one.

That's right.

There's more horses in that one.

Can you, can we still do long-term planning? Does it still matter for people to think long-term?

It matters more than ever. I was born in 1971. I was 8 years old when Stanislav Petrov saved the world. Do you know this story? He was a Russian Air Force commander and he saw what looked like US missiles coming into Russia and he had to make the decision whether to fire back or not. And he decided not to and he saved the world. I never thought I would see 30. You know, I'm 55 years old now. I bless my northern mother who was always like, don't get into debt, pay off your credit card every month. Real estate is the real wealth because she made me invest in my education and she made me save for the future. And here is the future completely unexpectedly. We didn't have a nuclear war. We came so close so many times and we didn't have it. Long We are coming so close to climate catastrophe, but the only way forward is to keep educating your kids, is to keep building sustainable homes, is to drive electric vehicles, is to put solar everywhere, is to plan for the future. If we all die, doesn't matter. If we don't all die, we had a plan.

And, but we first have to have that plan for For us not to all die?

We have to have a vision of the world we want. You know, we keep coming back to this. This is what I see in the founders that I back. We have to believe that this world is not perfectible, but that it's improvable. That not so many people have to die of preventable illnesses, that not so many people have to miss out on the education that would transform their lives. We have to believe that the world can be better.

Do you invest in climate companies?

I don't. Well, I do. I did out of my personal money before I started Generation Ship. And actually one of the best completely impulsive investments that I've made is in a company called Critical Loop, which I invested in because I was starstruck by the founder, Bala Ramamurthy, who came through Alchemist Accelerator, which I was running, and he's He was at SpaceX. He was the director of human spaceflight safety. I just love Bala. He's incredible. And he came in with this prospect. It was going to be a bank. It was going to be a climate thing. And I'm like, I have no idea what you're talking about, but here, take some money. And what he's ended up doing is building mobile batteries that charge at solar stations and then sit in factories and at airports and provide backup that isn't a generator. And that company's Doing incredibly well. Again, he had a vision of the future that I couldn't even follow, but it sounded really cool. So I threw some money in. And the reason I didn't build Generation Ship around climate tech is just, I don't know enough about it. I know so much about infrastructure. I've forgotten more about infrastructure than I've ever known about climate tech. And so that's my weirdness. That's my, that's the thing that makes me me. And when I Got my 100 rejections and edited myself down to, this is the thing that's interesting about me. What's interesting about me is that I find infrastructure software fascinating.

Well, at least I know you're passionate about it, and you're telling our kids to, you know, we need to all—

Oh, my kids absolutely hate AI. One of them wants to be a film director. The other one wants to be a firefighter.

What are you telling your kids to get themselves prepared? AI is killing the career ladder that we grew up with.

Yeah, I went through a prehistoric version of this when I realized that being an English professor wasn't going to play out. It was like, given the uncertainty of the markets, given the uncertainty of geopolitics, what do I have? What makes me weird that I can bet on, that I can use as a commodity in this market? And I think that's happening to more and more people. And so, what I tell my kids is what I tell my founders, find that weirdness, find the thing that is completely unique about you, because all of the The grinding stuff, LLMs can do that now. Like all of the just repetitive, boring work, that's work for computers. It's not work for humans. Work for humans is caring about stuff. It's making intuitive lateral leaps. It's taking care of things, you know, things like nursing and teaching and caring for animals. It's really hard for computers to do those. Having novel thoughts, almost by definition, LLMs can't have novel thoughts. They can only statistically reproduce the most likely thing that happened out of the corpus. So, where do humans get our advantage? In being weird and messy and organic and embodied and physical. The thing is to find the thing that you are uniquely qualified to do. It's to find where you fit in this ecosystem.

How do anyone find how they fit in this new world?

Just get a lot of rejections.

Same as being a founder then. Just get your first 100 rejections.

Yeah. Try lots of things and see what fits.

I have a traditional closing question for all my guests.

So excited about this one.

If you could recommend a book, I can see you have a lot. Recommend us a book. That can change our lives or that has changed yours?

So last year I finally read Origin of Species in its entirety. I had tried to read it when I was about 15, cuz I was that kind of weird kid and it's dense. It was a really challenging read when I was 15. Coming back to it after all that time was like, I visited La Sagrada Familia, the cathedral in Barcelona when it was under construction and I visited it again when the windows had gone in. It was like that coming back to Origin of Species. It was like, oh my God, this is, this is not a construction site. This is a, I think it may be the best book written in English. So, so many people get this wrong because I think more people allude to Darwin than have ever actually read the book. But he spent the first half of his life exploring and collecting and surveying the space. And then he moved to his house in Downe and just studied and had this huge network of correspondents and thought about his collections for the next 40 years, which is super relatable because I've done the same thing. And Darwin was really weird and he was curmudgeonly and hypochondriac. But what he did was he looked at all of this new evidence, the fossil record, the finches that he'd observed in the Galápagos, all of the people he was corresponding with. You know, he became a member of the fancy pigeon breeders association in London, which was this totally working class thing. And he felt it was a huge honor that he He was accepted among these, these people who were breeding the fancy pigeons. And he looked at all of this data and said to himself, how do variations within a species, do they have some relationship to the existence of different species? Given that geological time is not 4,000 years as the Bible tells us, but is millions and millions and billions of years. Given enough time, could these variations Separate out into species. It is a work of such imagination and discipline. And because Darwin had this network of correspondents all over the world, it's also a work of great empathy because he's anticipating the objections that will come from his scientific colleagues and, and talking them through each one. It's also incredibly moving because while he was writing this, he and his wife lost their favorite daughter at 5 years old. And You know, it's the mid-Victorian era. You could be expected to fall back on faith in Jesus and in heaven. But what Darwin was finding was that you don't actually need a creator God to explain the multiplicity and the beauty of the Earth. You can actually use scientific principles to explain this. And I want to finish on this note that everybody associates the phrase the survival of the fittest with Darwin and think that He's this like Hobbesian philosopher about nature red in tooth and claw. He didn't even come up with that phrase. That's Herbert Spencer's phrase that Alfred Wallace fell in love with and persuaded Darwin to take on. Darwin's own phrase is natural selection. And even when you're talking about the survival of the fittest, it's easy to think about like gladiators and MMA fighters and gym bros. But that's not the kind of fitness he's talking about. He's talking about fitness for purpose. He's talking about how organisms fit into an ecosystem. And so again, you end up with this authoritarian world versus this world of multiplicity and beauty and variety. And Darwin's world is the world of multiplicity and variety. He doesn't fall back on a father creator God when he loses his precious daughter. He looks at the world as it really is. He looks at how it changes very slowly over time. He looks at how organisms find their place. In these very complex and vibrant ecosystems. And that's where he finds comfort. It's a novel with the impact of a Middlemarch or a Ulysses, and it's also a landmark work of scientific literature, and it's absolutely beautiful piece of writing.

You know, my heart is pounding when you were talking about in the last 5 minutes, and I feel this book pretty much represents your investment thesis as well. Right. It's the multiplicity of the human society. Find your 100 rejections. That's the takeaway. Thank you. See you next time.