What Surprised Us Most In 2025

I think perhaps the thing that most surprised me is the extent to which I feel like the AI economy stabilized. We have, like, the model layer companies and the application layer companies. And the infrastructure layer companies seems like everyone is gonna make a a lot of money, and there's kind of, like, a relative playbook for how to build an AI native company on top of the models.

Many episodes ago, we talked about how it was felt easier than ever to pivot and find a startup idea, because if you could just survive, maybe you could just wait a few months, there was likely gonna be some like big announcement that would completely make a new set of ideas possible. And so, like, finding ideas is sort of returning to sort of normal levels of difficulty.

Welcome back to another episode of the Light Cone. Today, we're talking about the most surprising things that we saw this year in 2025. Diana, you found a pretty crazy one. It's sort of a changing of the guard almost in who is the preferred LLM at YC during the YC batch.

Came out a bit more than OpenAI, which who'd have thought? I think when we started this podcast series back then, OpenAI was, like, 90 plus percent. And now, Anthropic. Who would have thought?

Yeah. And, you know, they've been hovering around, like, 25% for most of, like, twenty twenty twenty four and early twenty twenty five. And then only even in the last three to six months did this sort of changing of the guard actually happen.

They had this hockey stick with the with the growth of over 52%.

Why do you think that is?

I think there's a couple of things in terms of the text tag selection. I think as we've seen this year, there's been a lot of wins in terms of vibe coding tools that are getting built out out there. And coding agents are so many categories that this ended up being a bigger problem space that actually is creating a lot of value. And it turns out the model that performs the best at it is actually the models from Anthropic. And I think that's not by accident.

I think from the hearing the conversation we had with Tom Brown not too long ago, he came and spoke because that was one of their internal evals. They, on purpose, made them their North Star, and you can see it in the model taste as a result of what's the best choice of model for a lot of founders building products is Anthropic.

The vast majority of the use cases people are using it for though is not coding. So I wonder if there's like a bleed through effect where people are using Claude for their personal coding, and then as a result, they're more likely to choose it for their application, even if their application is not doing coding at all.

Because you'd be very familiar with, like, the personality of Claude Opus or whatever they're choosing. Yeah. Sonnet, I suppose.

How about Gemini? How's Gemini doing in those rankings?

Gemini's also pretty much has been climbing up pretty pretty high. I think last year was probably single digit percent or even, like, 3%. And now for winter twenty six, it's about 23%. And we personally been using also a lot of Gemini three point o, and we've been impressed with the with the quality of it. I think it's really, really working.

I mean, they have all different personalities, don't they? Too. Yeah.

It's it's kind of the classic where OpenAI sort of has the black cat energy. And almost like Anthropic is kind of more the happy go lucky. I mean, more very helpful golden retriever. At least that's what I feel when I talk to them.

And how about Gemini?

It's kinda like in between.

Harj, you prefer Gemini actually.

Yeah. I switched to Gemini this year as my just go to model. I think even before 2.5 Pro came out and just seemed better at reasoning for me. It was just like the increasingly I replaced my Google searches with Gemini, and I just sort of trusted that Google's I think like the Groundings API and its ability to actually, like, use the Google index to give you, like, real time information correctly. I just found it was better than Personally, found it was better than all the other tools for that, and it was better than Perplexity on it too.

Like, Perplexity would be fast, but not always accurate, and Gemini was not quite as fast as Perplexity, but was always pretty accurate if I asked it about something that happened today.

Even if you use Gemini as the reasoning engine in Perplexity.

I have not done

that. Interesting. Yeah. I just It's hard to know, like, how much of it is the tooling and how much of it is, like, the base LLM.

That's fair.

Yeah. I mean, what are your guys' tools of choice? I haven't switched off of ChatGPT. I mean, I find the memory very sticky. It knows me, it knows my personality, it knows the things that I think about, and so I'll use perplexity for fast web searches or things that, you know, I know is like a research task, because I think Chechiuti is still like a little bit of a step behind for searching the web.

I don't know. I think memory is turning into an actual moat for, like, that consumer experience. And I don't expect Gemini to ever have the personality that I would expect from Chattypie T. It just feels like a different, like, entity. You know?

The thing I'm still surprised about is why there just aren't more consumer apps around all the various things we do. Like if I think back, one of the big changes for me this year was just the amount of prompting and context engineering I do for my life. We bought a house recently and the whole thing, I just had a really long running chat GPT conversation, stuffing it full of context of every inspection report or wanting it to be level the playing field between me and the realtor to understand all the dynamics and things that are going on. And it just feels like there should be an app for that.

But simultaneously, I'm sure you took the PDFs and just, like, dropped them into Gemini and said, well, summarize and tell me what's important for me.

I guess I worry about I worried about I still don't trust the models enough to be accurate without lots of prompting, and it's a high value transaction, so you don't want to like, get incorrect data out of it, so I still feel like you need to put in the work, and it feels like there should still be apps that just do all the work for you.

Did you see Karpathy release, like, sort of a LLM arena of a sort? Which, I mean, I do by, like, hand right now using tabs. It's like, you have Claude open, you have Gemini open, you have Chachi Pitea open, and you give it the same task. And then you take the output from each, and then I usually go to Claude at that point. And I'm like, alright, Claude.

This is what the other one said. What do you think? And check each other's work.

I actually think that that particular behavior at the consumer that level that we're doing, startups are doing as well. They are actually arbitraging a lot of the models. I had some conversations with a number of founders where before they might have been loyalists to, let's say, OpenAI models or Anthropic, And I just had some conversations recently with them, and these are founders that are running larger companies like series b level type of companies. With AI, they're actually abstracting all that away and building this orchestration layer where perhaps as each new model release comes out, they can swap them in and out, or they can use specific models that are better at certain things for just that. For example, I heard from the startup, they use Gemini three to do the context engineering, which they actually then fed into OpenAI to execute it.

And they keep swapping it as new models come up, and the winner for each category or type of agent work is different. And, ultimately, they can do this because it it is all grounded based on the evals. And the evals are all proprietary to them because they they're a vertical AI agent, and they just work in a very regulated industry, and they have this dataset that just works the best for them. I think this is the new normal right now where people are expecting, yeah, the it's cool that the model companies, they're spending all this money and making intelligence faster and better, and we can all benefit. Let's just do the best.

It's almost like the era of Intel and AMD with new architecture would come up. People could just swap them. Right?

Yeah. It feels like the highest level that angst around where's the value going to accrue. Is it gonna go to the model companies or, like, the application layer, I. E, the startups, feels like that ebbs and flows in either direction a little bit throughout the year to me. Like I feel there are moments where Claude Code, amazing launch, and it was like, oh, okay, the model companies are actually gonna play out the application layer.

But then to me at least, it's all vibe space. Gemini Surge, especially over the last few months, feels like it returns us to a world of where exactly that. The models are all essentially commoditizing each other and it's just the application layer and the startups are set up to have another fantastic year if that continues.

I'm curious what you think, Jared, with a lot of, perhaps the negative comments on Twitter around, is this a bit of a bubble, AI bubble?

Yeah. When I talk to undergrads, this is like a common question that I get is like, oh, like, I heard it's a big AI bubble because, like, there's all this, like, crazy round tripping going between NVIDIA and OpenAI, and like, is it No. This is great for you. Is is is it all fake?

Yeah. No. This is fantastic. Right? Like, people look at the telecom bubble, it's like, there's just, you know, billions of dollars, like, tens of billions, hundreds of billions just, like, sort of sitting in a bunch of telecom back in, like, the, you know, nineties.

Actually, that's why YouTube was able to exist. Right? Like, if you just have a whole bunch of extra bandwidth that isn't being used and is relatively cheap, the cost is low enough for, like, something like YouTube to exist. Like, if there wasn't a glut of telecom, then like, maybe YouTube would have happened and just would have happened later. And then that isn't that, like, sort of what we're talking about here?

Like, how do we we have to accelerate. Right? We have the age of intelligence, the rocks can talk, they can think, and they can do work, and you just have to zap them more. And you get like smarter and smarter stuff at this point. I think the argument to college students is actually like, because there will be a glut, there is an opportunity for you.

And if there was not a glut, then there wouldn't be as much competition. The prices would be higher. The margins lower in the stack would be higher. Right? And then, you know, what's one of the big stories this year?

Like, Nvidia suddenly is on the outs. Like, I think their stock is today is like around 1 seventies or something. You know, I think I'm still at long term buy and hold honestly, but for the moment, people are like, oh, well, Gemini's so good, and all the you know, nobody seems to be NVIDIA only now, and everyone's buying AMD, and everyone's you know, and TPUs are working. So, you know, at the moment, it looks like there's, you know what does that mean? Like, there's competition, and it means that there will be more compute, not less.

And then that means that probably a little bit better things for all of the big LLM companies, like sort of the, you know, the AI labs, they get a little bit of power. But, you know, they too are in competition with one another. So then what does that mean? Well, it's, you know, go up another level in this stack. Right?

Like, as long as there are a great many AI labs that are in deep competition with one another, then that's even better for that college student who's about to start a company at the application level.

Yeah. I think that's exactly right. It's like, people are asking this question like, is it a bubble? That's maybe a question that's really relevant if you're like the equivalent of like Comcast. Like, if you're NVIDIA, that's a very relevant question.

Like, oh, are people overbuilding GPU capacity? But like, the college students, they're not Comcast. They're actually like YouTube. If you're doing a startup in in your dorm room, it's like the AA equivalent of like YouTube. And like, content doesn't really matter that much.

Maybe NVIDIA stock will go down next year. I don't know. But, like, even if it does, that doesn't actually mean that it's, a bad time to be working on an AI startup.

Yeah. It's what Zuck said in a podcast earlier this year, think. Right? It's like Meta may end up over investing a significant amount in the CapEx and infrastructure. But essentially have to the big companies have to do it because they can't just sit on the sidelines.

And in the case demand falls off a cliff for some reason, it's their CapEx, not the startup's CapEx. And there's still gonna be tons of infrastructure and ideas to still continue building.

There was this book written by this economist called Carlota Perez, who studied a lot of tech trends, and it studies a lot of technology revolutions. And it summarizes us there's really two phases. There's the phase of installation, which is where a lot of the very heavy CapEx investment come in. And then there's the deployment phase where really ripples where it rips and then everything explodes in terms of abundance. And during the initial phase of installation is where it feels like a bubble.

There's a bit of a frenzy because it starts first with a there's this new technology that's amazing, which happened with the Chattypity moment in 2023. Everyone got super excited about the tech, and then everyone's got super hyped and got into investing into a lot of the infrastructure with buying a lot of GPUs and all the giant gigawatt data center build out. And then people say, but what is the demand? What are gonna be all the applications to be built out? I think right now we're in that transition, which is actually really good news for startup founders because they are not involved into the building the data centers, but they're going to build the next generation of applications in the deployment phase when it really proliferates.

And what happened, just going back to the analogy with with the era of the Internet, before the 2,000, there was a lot of heavy CapEx investment into the telcos. Right? Those were giant projects that college students wouldn't be involved, but they were very heavily invested. And in some cases, we're overinvested. I mean, there's a whole thing with dark fiber and some pipes that are not used, and that's fine.

The Internet ended up being still a giant economic driver. And what that means is startups like the future Facebook or the future Google are yet to be started because those come in in the deployment phase. Because right now, think things things are still getting built up. I I do think the foundation lab companies and GPUs very much are falling into the bucket of infrastructure.

Yeah. I mean, it's interesting to watch how this stuff is evolving a little bit. So do you remember summer twenty four, there was a company called Star Cloud that came out, and was one of the first to come out and say, we're going to make data centers in space. And what was the reaction when, you know, people laughed

at them. Yeah. On the internet. Yes. They said that's the stupidest idea ever.

You know,

I guess eighteen months later, suddenly Google's doing it, Elon's doing Elon's it.

Is in every interview now, apparently. Is that right? It seems to be like his top talking point.

Yeah. And so, I mean, why is that? Like, I feel like one of the aspects is that, part of the infrastructure build out right now that's so intense is, we literally don't have power generation. Yeah. Boom Supersonic, instead of making supersonic jets right now, is on this good quest to create enough power for a bunch of these AI data centers that are being built right now.

They use jet engines, and even those, like, are so you know, the supply chain for jet engines to generate power for data centers is so backed up that, you know, you would have had to have ordered these things, you know, two or three years ago just to even have it in two or three years from now. You know, these constraints end up, like, influencing, like, fairly directly what the giant tech companies need to do to win the game three or five years out. Like, suddenly, there's not enough land. You know, in America, we can't build. The regulations are too high.

In California, we have CEQUA, which is totally abused by the environmental lobby to stop all innovation, and building housing, by the way. We just don't have enough terrestrially, like, to just do the things that society sort of needs right now. So, you know, the escape valve is like, actually, let's just do it in space.

Yeah. Come to think about it, we we kind of have the trifecta of YC companies that are solving the data center build out problem. Well, you need fusion energy. Yeah. Yeah.

Well, we have the company that's solving the no land problem by building the data centers in space. We have Boom and Helion, which are solving that we don't have enough energy problem.

I just funded a space fusion company that just graduated called Zephyr Fusion.

That's a cool one.

And they actually had a great seed round out of demo day. They're in their forties. They're national lab engineers who, their entire careers, they were building, you know, tokamak's and fusion energy. And they came into the lab one day. They looked at the physics.

They, you know, looked at the math and the models, and they said, you know what? If we did this in space, it would actually pencil. And so that's they're on, like, this sort of grand next five, ten year quest to actually manifest it, to actually create it in space, because the equations say that it is possible. And if they do it, it's actually the only path to gigawatts of energy up there in space. So, you know, it might be, you know, an even more perfect trifecta shortly.

Something else I feel like happened over the course of this year is the interest in starting model companies. Like, I guess that maybe at both ends, there's, like, the people who can raise the capital to go and actually try and build a head on competitor to OpenAI, which are very, very few. Like, maybe Ilio with SSI. But then more so within YC, people are trying to build like smaller models. I've certainly had more of those in the last few batches than before, like whether it's sort of like models run on edge devices or maybe like a voice model specific to a particular language.

And I'm curious to see if that trend continues going back to this early era of YC actually. We sort of saw the exposure of just startups being created and maybe especially SaaS startups. Partly what fed that was just knowledge about startups became more dispersed. There wasn't like canon of library information on the internet about like how to start a startup, how to build software, and then over like a decade that just became more commonplace and that just exploded like society's knowledge of startups and how to build things. And feels like maybe we're going through that moment in sort of the AI research meets actually building things.

With training models,

I think we are absolutely going through that right now, yes. Where it's going from being a very rare skill set to a more common one.

Because like OpenAI a decade ago was like a rare like, you need you need like a a unique combination of skills. Right? You need like your researcher brain, your sort of like engineering brain, maybe like your sort of finance business brain,

or Wait. Wait. Wait. So did you just describe Ilya, Greg, and Sam?

You got it. Yeah. That was like a rare team. Right? There just wasn't that configuration of team around very much.

And now, a decade later, there's a plethora of people who have the research background, the engineering background, the startup capital raising background, or at can be taught how to do all of that kind of stuff, and I'm curious if that would just mean it would just seem more applied AI company starting, maybe there'll be even more models to choose from for all the very specific tasks.

I think so. I think the other thing that's even contributing and making this an even bigger snowball is because of Aural. I think there's all these new open source models that people are doing the fine tune on top of it with a particular Aural environment and task. So it is very possible that you can create the best domain specific, let's say, health care model trained on a generic open source model by just doing fine tuning on it and doing RL. It beats the regular big model.

Actually, I've heard and seen a number of startups where their domain specific model beats OpenAI, let's say, health care. There's this particular YC startup that told me that they collected the best dataset for for health care, and they ended up performing better than OpenAI in a lot of the benchmarks for for health care with only 8,000,000,000 parameters.

I guess what's funny is you do need to have a post training infrastructure. You know, we've also had YC companies where they had something that beat OpenAI, you know, GPT 3.5, and they were doing fine tuning with RL. But then, yeah, GPT 4.5, and then 5.1 came out, and, you know, basically blew their fine tuning out of the water.

You have to keep going. Yeah.

Yeah. You gotta keep going. Yeah. I mean, you actually have to continue to get to the edge. Anything else that really sort of stood out from this past year that jumps out to you?

It's funny. We started the year with one of our episodes that got a lot of views around vibe coding. I think we were talking about it more as observing a behavior that was happening from our founders. And I was surprised to see that this became like a giant category. There's lots of companies that are winning.

I mean, we have Replit. There's Emergence. There's a bunch of them.

Varun Mohan had gone over to Google. He released Anti Gravity. And did you guys see the video? Actually, I'm sort of curious whether they actually used Nano Banana or any of these video gen things, because it's like a little too perfect, but Google has the budget to do the high production value video. But it's, you know, Varun at the keyboard, and then, you know, Sergei is like right behind him.

So I was like, this is very cinematic. Anyway, I think Sundar was, you know, also not only talking about space

Data centers. Data centers.

He was also talking about vibe coding. And I knew that I was a little bit trolling back, but knowing what we know I mean, yes, vibe coding is not, you know, completely usable and trustable for, you know, a 100% of your coding, period. Like, this you know, it is not true that you can, like, ship a 100 a 100% solid production code today as of 2020 and the 2025.

Yeah. I was thinking about things that surprised me in 2025, and I think perhaps the thing that most surprised me is the extent to which I feel like the AI economy stabilized. Like, I feel like when we did this episode at the 2024, it felt like we were still in the middle of a period of incredibly rapid change where the ground was shifting under our feet and, like, nobody knew when the other shoe might drop and, like, what exactly was gonna happen with startups and AI and the economy. Now I feel like we've kind of settled into, like, a fairly stable AI economy where we have, like, the model layer companies and the application layer companies and seem the infrastructure layer companies seems like everyone is gonna make a a lot of money, and there's kind of, like, a relative playbook for how to build an AI native company on top of the models. I feel like things really kind of matured in that way.

Which feels is all downstream of like the models themselves have incrementally improved this year, but there haven't been like major steps forward that have shaken everything up, which is, has a knock on effect. Many episodes ago we talked about how it was, felt easier than ever to pivot and find a startup idea, because if you could just survive, if you could just wait a few months, there was likely gonna be some big announcement that would completely make a new set of ideas possible and create more opportunities to build things. It certainly feels like that has slowed down. And so, like, finding ideas is sort of returning to sort of normal levels of difficulty in my experience of office hours. I agree.

I'll tell you what's not a surprise. Do you remember that report AI 2027 where it was just sort of like this doomer piece that said like, oh, well, society is gonna start falling apart in 2027. But, you know, at some point, they quietly revised it to say that it wasn't 2027, but they kept the title. Maybe it's not a surprise. Like, I was always a little bit of a skeptic of like this fast takeoff argument, because even with the scaling law, it is log linear.

So it is slower. It requires like 10 x more compute, and it's still sort of, you know, topping out. Right? And that's one form of good news. Another form of it's weird to call this good news, but human beings don't like change.

In our previous episode where we sort of blew up that MIT report that said that, you know, 98% or 90% of enterprise AI projects fail. Well, it turns out that 90% of enterprises don't know how to do, you know, IT, let alone AI. It's weird to say that that's a good thing, but in the context of Fast Takeoff, like, that is a real break on the ability of this new, really insane technology from actually permeating society. I love to accelerate, but like, it's weird to say like, oh, well, actually, in this case, maybe that's a good thing. Right?

Like, it is a shockingly powerful technology. But, you know, between being log linear scaling, and human beings really don't like change, like, organizationally speaking, society will absorb this technology. Everyone will have enough time to sort of process it. Like, culture will catch up. Governments will be able to respond to it, not in like a frantic s b ten forty seven sort of like, you know, let's stop all the compute past ten to the twenty sixth.

Right? Like, just these knee jerk responses to technology. We're excited about the ARC AGI Prize is, you know, gonna come in and do the winter twenty six batch as a nonprofit. The funny thing about that is like, yeah, maybe there's a team right now that is climbing the leaderboard of Arc AGI, and they're gonna accelerate this thing again.

Something that surprised me to relate to that with the startups is I remember around this time last year, were talking about how companies are getting to a million dollars ARR and raising series A's without hiring. Like, SimCase is not hiring anyone, just the founders maybe hiring one person, which just felt very unusual. I feel like a year on that hasn't translated into, okay, and then they went and hit like 10,000,000 ARR, or they scale without adding any more people too.

No, they turn around and start hiring like actual teams.

Yeah, post series A, it actually largely feels like the playbook is the same. And the companies might be smaller for the same amount of revenue, but it feels it's entirely because they hit the revenue so fast and they're still just bottlenecked on how long it takes to hire people versus they have demand for less people.

I still think there is, like, a you know, some effect, but it is not, like, open and shut. It is not like you don't have to hire executives anymore. I think there are, like there might be a case of two foie gras start ups, like one being Harvey and the other one being Open Evidence. Right? Harvey, the founders are incredible.

They were, you know, very early. And then there's this sort of idea of like, for VCs, you could just go down Sand Hill Road and like the fix is in. Like, you just sort of block out all of them. And then all the people, you know, there may maybe 30 people who could write checks of like ten to a hundred million dollars, and if you just sort of get all of their money, like, there's sort of no one who can actually come in and do the next series a, and then basically you're safe. Like, you have capital as a bludgeon is capital as a moat in that case.

Right? So, yeah, Harvey is interesting because, you know, Legora's coming fast for them. And obviously, we have some skin in the game on Legora, but we think that they have as good a shot at any.

I guess that's one trend that we saw in 2025 is that there was, like, a first wave of, like, AI head of companies like Harvey Who might

have wasted a lot of money on fine tuning, actually.

Totally. That, like, broke out really big in 2023 and kind did a victory lap that, you know, oh, we've won the the space, and now we're seeing a second wave of companies like Flagora and Giga. And it turns out that, like, oh, actually, like, it isn't so simple.

Yeah. The wood beneficiary of, you know, burning some non trivial double digit percentage of your capital stack on fine tuning that buys you no advantage is like, basically, the investors are the only winners there because they just own more of your company, you know?

Yeah. So at least as it relates to, like, the the hiring and team size, I feel like of the two camps, one being that AI is going make everything more efficient, you will need less people, and the other AI is going to reduce the cost of producing the time to produce things, and so then the expectations from your users and customers will just go up, and you'll need to keep hiring more people to satisfy the growing expectations. I feel like this year has been more in that second camp, and I think that is what's driving the fact that the companies are still just hiring as many people as they were pre AI. Is just like the bar for what their customers expect. And they're all in, you know, Lagora's racing with Harvey, Giga's racing with Sierra, like they're all still competing for the same set of customers.

And they still ultimately are bottlenecked on like people and like, I don't think anyone's bottlenecked on ideas, but they bottlenecked on, like, people who can execute really well. I don't I I that's, like, still it's Feels like an exciting phase.

I agree with you that, like, the era of the one person running a trillion dollar company is not here.

Not yet.

Yeah. But I think it's gonna trend that way eventually. That'll be a wild time. Maybe that's a prediction for

For next one?

2026. Yeah.

You think it's coming in

2026? It'll happen in 2026 either, honestly. I mean, I think you will have many stories of companies run by, you know, under a 100 people that are making hundreds of millions of dollars. Yep. So, I mean, Gamma was interesting to see.

Like, one of the biggest things that they said in their launch that I think is a very good trend is they said they got to a $100,000,000 in ARR with only 50 employees. Yep. So which is a very different it's, you know, such an inversion. Right? Like, normally, you have the big banner and the, like, little x thing, you know, image, and it's like, oh, yeah.

Like, we raised all this money, and look at all the people who work for us. It's a good trend to have the reverse flex, which is like, look at all this revenue, and look how few people work for us. Well, that's all we have time for this time. We just wanted to wish you a really happy holidays and happy New Year from all of us to you and yours. See you next time.