Why AI Labs With Unlimited GPUs Still Fail — Anjney Midha, AMP

But more, more concretely, there are so many AI labs today that have all the cash they need, they have all the compute they need, and they're still not able to ship anything SODA. And then you start seeing people leave and so on, and my diagnosis, it's, it's, it's the culture.

If you stop taking the actions that demonstrate the mission alignment to what you've said to your team and to your, the world matters to you, then your culture starts to fray.

Before we get into today's episode, I just have a small message for listeners. Thank you. We would not be able to bring you the AI engineering, science, and entertainment content that you so clearly want if you didn't choose to also click in and tune into our content.

We've been approached by sponsors on an almost daily basis, but fortunately, enough of you actually subscribe to us to keep all this sustainable without ads, and we wanna keep it that way. But I just have one favor to ask all of you.

The single most powerful, completely free thing you can do is to click that subscribe button. It's the only thing I'll ever ask of you, and it means absolutely everything to me and my team that works so hard to bring The In-Space to you each and every week.

If you do it, I promise you, we'll never stop working to make this show even better. Now let's get into it.

We're in Periodic Labs with Anj Midha, CEO, uh, founder of AMP. Welcome.

Thanks for having me. At Google, uh, if utiliz- I- so there's two types of utilization usually, right? That you're measuring in these clusters. One is node allocation, and then the other's MFU. So node a- utilization is usually just what percentage of cards in the data center are just, like, used, and that, if it's not at, like, 95%-

There's no excuse

... there's no excuse, right? Like, I think 95% at Google, which is where my co-founder, Seb, uh, came from, uh, he built the BorgX, Borg GQM scheduler at Google, and there I think 95% was considered an outage. So 96% node utilization is sh- should be standard.

And, and most single tenant clusters are not running at that. So that's one. And then MFU utilization should be, I would say the best in class today is somewhere between 60 and 70%. Uh, I, I think this is a leadership question, right?

Is, is there an... And, and u- f- fundamentally it's an alignment question, which is are the people who are funding the cluster and then deploying the cluster actually aligned? And sometimes theoretically they are, but in practice, the number of people in the chain, the supply chain between, like, the capital and all the way to whoever's managing the cluster and then whoever's measuring what the output is, are just so many, you know, degrees of separation away that, like, the, you know, the...

Have you ever heard that sort of r- you know, radian metaphor, which is at the beginning of, of, of an arc, if you have two arcs that are two lines that are just off by a few degrees, that-

It spreads out

... it spreads out, right? At scale. And I think what's happening is a lot of cluster implementations and infrastructure, a, a lot of frontier labs and other teams, that's what's happening, is they're, they, they initialize the plan, which is kind of like North, North Star with a team that wants to do good, but then they're required to scale so fast instead of iteratively that the wastage just compounds really fast at scale.

And so I, I think we know the answer, which is just do iterative bring ups, you know? Um, if you spend time with people who've been in the semiconductor industry or the DSN industry for a long time, this is not new, and I don't think AI should be an excuse.

Like, sure. S- something... What is new? Okay. We have a lot of new capabilities, but that doesn't mean just abandon common sense. Common sense should always be in fashion. You know? Like, AI scaling doesn't change the sh- in fact, if anything, AI scaling should be putting a premium on the value of common sense and infrastructure because the, the margin of error now is so much lower and the costs of wastage are so much higher.

Yeah.

And, and the cost of wastage, by the way, is not just economic. I'm obviously, I'm, I'm an investor or I'm an investor by background over the last few years. Now we're running an AI infrastructure business called, you know, AMP.

And I think that it's okay to say this time is different on the capabilities front. Like, we are genuinely getting capabilities at, of, of the, of a kind we haven't had before. That doesn't give you an excuse to say this time is different for everything,

Mm-hmm

... especially infrastructure. So look, I, I love the hacker mindset and the hustler mindset. Now, that's great for the startup mindset, but y- you remember this moment where Zuck went from saying, uh, "Move fast, break things" to, like, move-

Fast with stable infrastructure

... move fast with stable infrastructure. I think now we need to move fast with, like, responsible infrastructure.

Yeah.

They're going to say s- like, where is the impact, you know?

Yeah.

Um, there was a really in, in our class yesterday, Scott Nolan, who's the founder of General Matter, came by at Stanford to speak about energy bottlenecks, and he had a phenomenal idea. He said, "If you look at the marginal unit economics of compute per hour, let's call it like $4 an hour.

If you're having to bring up a new data center in a new community, why not just say we're gonna charge $4.50 an hour and that marginal impact or that marginal increase, we just literally take that and give it to the local community as cash?"

I can tell you as a customer of that compute, I would love that. I'd be happy to pay an additional 50 cents per hour at scale.

Wow. Yeah.

Because if that means the public benefit is so clear to the communities that the data centers are coming up in, I'm gonna feel like that compute is much more reliable. You know, up to 20% of all data centers this year in, uh, in the US my understanding is are at risk.

Of community backlash?

Uh, correct. Of, of not getting the community support they need to get brought up.

Wow. That's a huge number.

Yeah. Now we- I think we should dig into what that number is. I think it's a little bit of overstated. Um, these things can get over-reported, but it-

They, they don't just care about jobs. They care about all the other stuff around it, right? Like they, they care about power grid, they care about environments-

Power grid, permitting and so on. And, and imagine... I, I think if you said w- there's a new AI deal. If we're bringing up a data center in your community, we're actually going to reduce the cost of your electricity bill.

Okay, now we're talking.

Yeah.

Right? The, the community's going, "Okay. Now, now this is a deal. I feel like a partner in this."

Yeah.

Right now that's not happening.

Mm.

There will be audits There will be investigations. And when the, when the regulators come, I don't know when it's going to be, the folks who are moving fast and breaking things in the name of AI progress better be prepared.

That's certainly not how we're procuring compute. Or we- we're trying as much as we can to work with partners who have long-term track records, many of whom, by the way, are not like AI providers. I think this whole idea of NeoCloud being somehow this new category is a lot of marketing speak.

There are really good, reliable, trusted data center providers in America who've been around 20-plus years.

Yeah.

I love those folks. They know how to... Sure. Have, have they... Are they sponsoring happy hours at NeurIPS? No. Are they legibly bitter lesson build? No. Are they hanging out in m- you know, in like situationally aware parties?

No. But they're adults.

Yeah.

I trust them.

They can run LAN. They can run power-

They can run LAN, PowerShell. They have credit histories. We sit down, we have conversations. Many of them live in Silicon Valley. They've, you know, they've had to deal with the boom and bust cycles of the internet, and I love those folks.

You know, they, they are stable infrastructure partners and thinkers, and I think there's a lot of short-term thinking going on in the compute layer and it's gonna catch up to us. It, it, it's not gonna be, uh, good.

You talk about aligning incentives, um, and you know, I would think that aligning incentives means you have the full stack in one company, which is xAI and OpenAI, right? So you as a standalone infrastructure layer, why are you somehow more aligned to your portfolio companies than people who just own the whole thing?

In systems design, right, there's, there's two regimes of, uh, architecture, right? You have integration, and then you have pooling and utilization, right? So the-- or, or rather, the way to increase utilization often is you can do systems integration where you collapse a lot of process into one node, or you can pull out a process from a node and share that amongst various-- that resource amongst several different nodes.

Mm.

And so we see the, the AMP grid, which is, uh, what, what-- the, the system we're building here, which is basically a compute grid. You know, we're doing-- trying, trying to do for compute what the electric grid-

Power, yeah

... yeah, what the power grid did for electricity. It-- this is a pooling and utilization layer across clouds, and so we're actually the opposite of a full stack integration like approach.

Super horizontal.

Where, where, where it's much more horizontal, and it's, you know, it's multi-cloud, it's multi-silicon. Um, the goal is to try to make f-mega, you know, FLOPs flow like megawatts, and that is very hard to do today for many reasons.

Like, there's stranded pools of compute all over the place and, and there's no fungibility. And so f- right now we do it at, at the level of scheduling, and we often do it at the economic layer. But as we start to announce what we're working on, it's extraordinary, like how many folks are coming out of the woodworks and saying, "Hey, I'm actually working on a way to make compute fungible at this part of the stack and that part of the stack."

And as a grid, we'd like all of these folks to participate on the grid. So just, you know, people often ask me, "Andre, are you a NeoCloud?" And I go, "No, actually, NeoClouds are suppliers." Or sometimes they'll ask, "Are you a venture capital firm?"

I go, "No, actually, they are, they are demand like sort of off-takers of the grid." We see ourselves as what's called an independent system operator. So if you study the history of the electric grid, once it became legible to a lot of factories and industrial sort of participants that, hey, actually it turns out pooling is a good idea.

We should pool our generators instead of all having a h- a generator running at half capacity in our backyard. There was a need for an independent entity who could coordinate all these parties, uh, transmission line, uh, you know, power generation, uh, facilities, transmission lines, factories.

And that neutral coordination mechanism is very critical. In order-- If, if you study like the history of grids, the most enduring ones were those that never owned their own assets. They were ones that had, uh, or often started with long-term anchors who were uncorrelated sources of demand, a steel factory, a shoe mill or whatever, in a particular town who weren't competitive, where the steel factory want to spike up at night, the shoe mill wanted to spike up during the day.

So then you pool and you share, right? So each of you is guaranteed some base load, but then you, you kind of schedule your spikes to drive a peak utilization across the town. The gold standard, so to speak, historically, has been these utility companies like PJM Interconnect in the northeast of, of America, where they, over many, many years, became this inde-- what, what's called an ISO, an independent system operator of the grid.

So that's how we see ourselves.

Yeah.

Um, economically, that's what we are. From a technical perspective, we started at the scheduling layer because Seb and Mihai who, uh, run engineering here built that-

Mr. Scheduling.

They, they did that at Google.

Yeah.

Um, and, uh-

And you have infra shops from Discord as well.

I have some.

I, I don't know, I don't know if Discord is like the primary identity, but what-whatever, I'm just kinda-

No, Discord was-

... choosing a well-known name.

Well, I, I so I was running the developer platform there.

Yeah.

The internal infrastructure I was not responsible for. That was actually a guy by the name of Mark Smith, who was extraordinary. And yes, Discord did pool... So Discord is actually a counter example. I guess I had the chance to learn a lot about fully, full stack infra there because-

It's the same thing, yeah

... um, it's the, it's the other architecture, which is, um, Discord built its own WebRTC s- vo-voice and video infra.

Mm.

So Go- like Discord did not use-

For the calls, yeah.

Yeah. Did not-- for communication, Discord did not use third-party infra. It was all built in-house. And then the way you maximize utilization was you pool demand from the world's two hundred million plus monthly active gamers, right? Um, and so, so that's, that's how those stacks were, were constructed.

You know, again, in systems design, the two concepts that keep coming up over and over again are abstraction and, and composition, right? And-

Bundling and unbundling.

Bundling and unbundling, abstraction, composition, like verticalization and, and-

Horizontalization

... horizontalization. So in that sense, AMP is an independent system operator of the grid. We pool demand, uh, we pool supply from a number of partners we trust at about one point three gigawatt scale over four years, um, and then we pool demand from some of the world's best, you know, research labs and so on.

We're sitting at one, you know, Periodic Labs, who need extraordinary long-term demand. And the idea is that- You know, each of them is guaranteed base load on the grid, but they can spike up and down flexibly on, for compute, uh, it, with much shorter timelines as needed.

That, that was roughly the design of the program I came up with at a16z called Oxygen. The same-- That was the same design of the, the GQM, BorgX, Borg GQM implementation at Google that Mihai and Seb had built-

Mm.

Which was that how do you allow, uh, teams inside of Google, uh, on the internal infrastructure to be guaranteed capacity, uh, for their base workloads, but when they need to spike up on research, how could they ensure that that was sufficiently there?

And of course, the big innovation that was discov- not discovered, but i- kind of implemented in the space, this infra space maybe three, four years ago at Google was the idea of interruptible demand, right? Where you just queue up a bunch of jobs and through this like sort of credit system, um, there can be a bidding mechanism.

Like priorities or, yeah.

It's a dynamic prioritization-

Yeah

... basically, and, and jobs can get interrupted based on somebody else who's saying, "You know what? I have 10 tokens, 10 credits I wanna spend on this job." Another business, like team lead, research lead is like, "Genie three or whatever is only worth five, you know, credits and NanoBanana two is worth 10 credits, and so the NanoBanana job gets priority."

I, I, that's a, that's a made up example.

It's very real. Brain Marketplace was real.

Yeah.

And, uh, we've, we've covered this on the pod with David Luan who was-

Oh, great. Okay

... who was there.

Awesome.

Uh, and the, the criticism is that, well, actually sometimes you need central command to go all in on the thing. Uh, and actually sometimes capitalism via credits doesn't work. Not, not, this is not a criticism of AMP. I'm just saying like this is a thing that has been tried, uh, internally within Google and it led to Google missing GPT.

Like we structured ourself essentially very similarly to Google. We, we are structured as a holdings company. So you know, Alphabet holdings is Alphabet holdings, and then they've got these subsidiaries called Google and-

Other bets

... Other bets and so on. We've got, um, you know, AMP holdings and we've got our infrastructure business and then we've got a capital business called Foundry that incubates new frontier AI labs and invest in them as venture capital, um, like Periodic.

You know, we put a few hundred million dollars into Anthropic from our fund earlier this year. So wherever we feel like teams are making progress, especially researchers and so on who've pushed the frontier inside of existing labs like DeepMind, I find, you know, there comes a point where they feel misaligned with the dictatorship of Alphabet holdings and at that point sometimes the dictatorship doesn't want them anymore.

Mm.

And they're like, "Thank you. You've done your job here. You've kind of helped us through the zero to one phase and for whatever reason we're gonna deprioritize your amazing like omnimodel or whatever it is, and instead we're gonna prioritize coding."

And uh, that, I think that's a tragedy but I get it. Like they're, you know, Sergey and team are running their own business there. But that doesn't mean we should, the rest of us should sit around waiting for that progress to get unlocked for the rest of the world and humanity.

I mean, if, if you think about how much extraordinary research has happened inside of DeepMind over the last 10 years, I, I, I mean, Demis and Sergey and those guys did such a great job. But at the end of the day, so much of that has never seen the light of day, you know?

Oh they, they're like papers only but they never actually shifted to production or-

I mean what's worse is the paper is actually not even being published anymore 'cause there's a six-month embargo inside of DeepMind, right? Like we've heard about this where a paper comes out and then I think there's a six-month embargo window where if anybody on the business team says this could be interesting-

Yeah

... it's embargoed for life.

Exactly. So the stuff that gets published is the stuff that's not good enough.

There's an adverse selection problem basically, yeah. At this point, well-

It's, it's a co- common complaint at NeurIPS by the way that's like well why would I look at the papers that are the trash of GDM?

Again, I think it's a tragedy. I mean, I get it. They're running their business but the rest of the spa- I, I, I think there's negative externalities of research being hoarded and so that's, there's a market failure.

Mm.

And, and somebody needs to unlock that research and we can't do it on our own. We only have 1.2 gigawatts of compute. That's nothing. That's about $40 billion of cloud spend. Um, we're gonna need a lot.

By the way is that, uh, that's a new number I haven't, haven't come across that uh, that gigawatt number. That's huge.

Yeah and then to be clear, we haven't secured all of it be- that's how much demand we have started to secure. I think publicly we haven't actually confirmed how much we have for this year in order-

Where do you wanna get to?

I think the steady state would be that we have a base load pool-

Mm

... of 1.3 gigawatts at all times-

Yeah

... of base load capacity. For spike capacity right now my estimate is we need roughly six gigawatts over the next four years for all our teams to feel like they were able to keep moving the frontier whatever they're working on whether it's, uh, like superconductor discovery over here.

There's a new investment we're working on right now which is in the end-of-life prediction space in, in healthcare. It's extraordinary how much you can, you can give peop- You know this was actually my graduate school work. I went to grad school for bioinformatics at Stanford Med.

Yes.

And I know we-

Econ MCS Bio.

Oh so my, I, I, I was this really weird cat where like I was never satisfied with my major options. So at one point I was an econ major then I was a CS major then I was a, a, a MCS major called mathematical computational science and they decided they were gonna end that major.

So I took all that coursework and I applied it to grad school, my graduate degree in bioinformatics, uh, which was the master's program and then I thought I was gonna do a PhD. I never ended up doing it.

I got, I, I, I dropped out and went to work at Kleiner but I was lucky enough to apprentice with this professor at, uh, Stanford Med. His name is Nigam Shah and he was working on end-of-life prediction. Stanford is one of the only research facilities in America that has a longitudinal patient data set that's larger at scale.

I think it's at least 12 million patient lives. The only larger data set is at the VA, the Veterans Affairs, you know, of America and to do research like do any deep learning and so on on that data set, it was called the Stride data set at that time, you had to be a Stanford Med school affiliate which is why I went and enrolled in the bioinformatics department.

Wow.

Um, end of deep learning was early. Nigam Shah had the visibility like the, the vision to see that like you could do end-of-life prediction to help palliative care. In, you know, in America the like over 30% of all Medicare, Medicaid spend at least at that time was spent on end-of-life care.

And what's, you know, we, we grew up in Asia so we all, yeah at least I, I won't speak for you but I have A very different relationship with death than I find folks who grew up in America do.

In America, spiritually and culturally, especially in Western societies where Christianity, the Christian tradition sort of frames death as this terminal point. Um, there's often a judgment day and so on. The, the way we view death is with a finality.

In Indian culture, in Hindu culture, you know, death is one-

I was raised Buddhist as well.

You're a Buddhist, yeah. So it, it's one, it's one step in a journey of many lives, right? And so, uh, I grew up in this city called Chennai in the south of India, and when people die, you dance on the street.

You know? There, there's like a procession where your, your body is carried to, to be cremated, and your family like, like celebrates, and there's drums and so on. It's this huge thing. And, uh, it, it's because the idea is that you, you're gonna be reincarnated.

You know, you've been liberated from the responsibilities of this life, and now you're onto your next. It's a new advent- It's like going off to a new college or whatever, right? And so it was so alien to me when I got here as an undergrad that it, the medical system works backwards from that assumption that we have to view death as this terminal thing and delay it, postpone it.

It's a bad thing. And so at the time, clinical decis- decision support in, in the United States was this very primitive field. Even to this day, physicians in the United States often will tell you when you have a terminal disease, "This is your...

We, we've diagnosed you," which is great. Our, our ability to diagnose here is extraordinary. "You have somewhere between six months to six years to live." What do you do with that information?

Yeah.

It, the error bars are so high that, that then you, you, you... W- In times of uncertainty, we default to culture, and when the culture is, "Let's... This is a bad thing, I've gotta prolong my life," then you start doing things like...

A- A- And just to, just sort of from a systems perspective, what's going on there is physicians often feel like they need to provide such high error bars because there's always some uncertainty in end-of-life diagnosis, and if you provide the wrong diagnosis or recommendation to your patient, you can be sued for medical malpractice.

Mm.

And then your license can be taken away. It can be catastrophic for your career. In contrast, if in countries where that's not the case, what you often observe is that patients... Like, physicians are quite prescriptive with their recommendation.

They say, "Hey, this is your condition. The literature says that you probably have this much time on Earth left. My expert opinion is that you are an outlier," or whatever, and they try to be more prescriptive. And that empowers a patient, right?

Because that patient can say, "I trust my doctor. They said on average, I have six months to live, but if I do these things, I may have a shot because of my particular predispositions or, or my genetic history or whatever."

And that empowers you to go about your life in a actually more scientific way than leaning on religion, culture, spirituality, and so on. In contrast, here, because of that medical malpractice sort of thing looming over your, your, your head, a physician never gives you a clear recommendation.

Right.

So instead you, you say, "Okay, Doc, well, let's try it all." And then you start a whole regime of drugs and therapies, and then you often spend weeks and weeks in the hospital, and that deteriorates your quality of life.

And when that deteriorates your quality of life, like, you instead, instead of spending your last few days doing the things you love with your family, you're spending it on a hospital bed, and that ends up being 30% of Medica- Medicare and Medicaid.

So it, it's worse for the patients. The doctors feel terrible. The American taxpayer is paying a huge amount of money. And so this is why Nigam Shah, who was this professor at Stanford, said, "Anj, if there's..." I, I kind of sat down with him.

I was this young... I'd, you know, I was 21, and I, I was like, "I, I wanna work on a big problem." He's like, "The big problem is end-of-life care." And so we tried to do deep learning to say...

Uh, uh, so we start trying to run deep learning on these stride patient data sets to say, "Could you have an AI system make a recommendation that is orders of magnitude more precise about how much time you have left once you've been diagnosed with a terminal condition than a human?"

And then if we can get that precision to be high enough, then you can empower the patient. And it turns out the tech works. Like it's, once you get the data set, like RL works. Honestly, even regression models work.

You don't need to get that fancy. At the time, we were just trying, doing like very simple neural nets.

Simple things. Yeah.

Today, uh, what we can do with RL is extraordinary. The re- problem remains w- then and now is regulatory because you actually can't shift the burden of the wrong f- clinical diagnoses from the physician to the AI system.

And so at that time, I got quite disillusioned 10 years ago for, uh, 12 years ago where, 'cause I felt I just didn't have the resources to influence regulation. Today, I'm very lucky. I'm in a different place. You know, I've, I'm a lot older, and so I've been spending a lot of time on my next incubation, which is how can we unlock the, like, patient empowerment by training AI models to do end-of-life prediction much, with much more precision and, and ac-

Oh, wow. You're still focused on this the whole time.

The... I, I, I haven't been able to get, uh, this out of my mind a single day for the last 14 years. Th- This is the hill I want, I would like to die on. This too, I would say.

You know what? I actually, I prefer not to die.

Yeah, exactly.

Uh, but, but I, I, I think two bipartisan issues, I think two issues that should be bi- bipartisan in America are how do we empower patients to make the right clinical decisions at the end of their life such that we're reducing the taxpayer burden with science?

It's just good old science, and AI can help here. And the second is, you know, net positive data centers 'Cause I think that's the biggest critical bottleneck on training en- good enough AI models to help people at the end of their life.

So, so there's sort of two sides of the, of the same scaling bottleneck curve, but th- th- those two, uh, you know, we formed AMP as a public benefit corporation. My wife and I, who you've met, you know, you've met Viv.

Yeah.

Um, her passion is, is education. Uh, you know, her family is a long line of educators and so on, and ph- uh, physicists. And so this, this class is my attempt to stopping the black sheep of the family and be a, an educator.

But if I'm not educating, the thing I would be doing is working, you know, on, on these two problems, whether on the political spectrum or as a researcher back at, at, in some lab. Um, and my hope is if anyone's listening to this podcast, if, you know, if, if they're passionate about either of those two topics, I'd love to hear from them.

We, we sh- we can share the contact in the show notes, but, uh, we're looking for people to join both of those missions on the, on the political side as well as on the medical side, on the research side.

You know, you said, uh, this is a s- a discipline that you want to form. You call it, it's called variously, variously called Frontier System. It's variously, variously called One Person Frontier Lab. Uh, what is the ideal name or shape of this?

Like, the what is the mission?

Of the class?

Uh, of the discipline that you're, I guess, exploring, right? Like, I, I... The, the class is called Frontier Systems.

Yeah.

But, like, for me, maybe one phrase is, like, uh, you- you're just anti-waste, right? Which is wasting, wasting GPUs, wasting in human and Medicare.

Yeah.

But, like, is there, is there a broader theme that I'm, that maybe you can encapsulate more succinctly?

Yeah, yeah. The, from an engineering perspective, it's very simple. It's output maxing.

Okay.

You know, it's the, it's the, it's the department of output maxing.

Make the most of what we have.

Exactly. I'm a huge believer in optimal outcomes. You know, I, I think both in America and other countries, uh, we are losing our appreciation for nuance, and this is the thing of... And AI is the same case, right?

Oh, the bitter lesson holds. Okay, fine. But that doesn't mean you just, like, throw 500 GB300, 500,000 GB300s at your, like, you know, suboptimal model scaling, and you waste a bunch of compute. It also doesn't mean that, you know, the most optimal is have, like, 50 different architectures, where there isn't enough standardization.

Like, o- one of the reasons Anthropic has had extraordinary sort of velocity is 'cause they picked the transform architecture and said, "This is simple. Let's double down on it," right? And now luckily there's enough investment going to the space that we can afford other architectures, but at the time, investment was just too fragmented into other architectures, so that un- arguably unlocked scaling.

So I think there's a philosophy. I, I think we all owe it to ourselves to do output maxing with a new capability called AI on a global level. I think if I was starting a new department at Stanford, depending on how fuzzy or technical I wanted to be, I'd probably call it the Department of Alignment.

Uh- You know, like-

It's an overloaded term.

But it is... But alignment really is a hard problem.

Mm.

And I think when you unlock it, full stack alignment is super hard in any organization, in any system. Like in a, in a venture capital firm, if you can have full stack alignment between your limited partners and your, the founders who are creating the value, and ultimately the public that owns the IPO stock, that is a gift that keeps giving.

And when you study the history of these systems, when they start off, they usually start out small scale, where the feedback loop is actually so tight that there's alignment.

Yeah.

And then the more you try to scale, the more division of labor happens, the more specialization happens, and at each step you add abstractions. And wherever there's an API interface, there's, like, loss. There's communication loss. And so I think a really cool thing would be for us to figure out is there a way for us to have our cake and eat it too as an engineering discipline?

Is there a way to actually scale up and scale out without losing any alignment, without lo- you know, lossy transmission?

Mean standards.

So standards is one way. The other way is you just have net new capabilities. So like sup... You know, what we're trying to do here is discover new superconductors. A room temperature superconductor would be a lossless transmission mechanism for energy.

I mean, we would have flying cars.

Yeah.

We are right within a few years of having a new room temperature superconductor. So I think those are the two. You either have to standardize on protocols or, or API specs that allow lossless communication, or you can come with a whole new capability that unlocks so much abundance, the standardization doesn't matter, 'cause you just unlock net new capacity.

Yeah.

Does th- so th- this is what I spend my days thinking about these days.

I mean, no, I think e- every infra person at, who wants scale and w- wants to output max does eventually end up thinking about this. Uh, we don't have time to go into it, but we have done an episode with SF Compute-

Oh, cool

... that is trying to standardize-

Yes

... the futures contract for compute.

Right. Right.

Uh, I don't, I don't know how that's going, by the way, but, like, at some point this, this will be part of it.

Oh, I think Evan is awesome, and- ... SF Compute is the kind of effort that I hope we can accelerate because what, what often happens is these exchanges are very hard to get, uh, th- it's hard to bootstrap them, right?

Because they often require... There, there's many inefficiencies between parties. There's trust boundary inefficiencies in infrastructure because you don't trust, you know, one part of the stack doesn't trust another part of the stack to give them visibility. There's capital markets inefficiencies, there's operational efficiencies.

So if you can inject like a, a single shock to the system of a ton of compute demand or supply, then y- you can accelerate, uh, th- these new flywheels. And so my hope is one day, you know, uh, or, or soon, if, if SF Compute needs extra capa- like, has excess capacity, they just hook it up to the grid, and they get flooded with demand from us.

And on the other side, if they have a ton of demand, but they don't have supply, they just again hook up to the grid, and it's a two-way protocol where they can just hook up to our capa- And I don't think we're too far from that.

You know, today our working implementation of it is mostly through a group of labs, um, universities and a few, like, uh, sort of trusted parties who are, who all- Feel like they're in alignment to borrow an over sort of used word.

Um, but our hope is to just have it be an open protocol that anyone can hook up to on-

Hook up for demand or hook up for supply? In-

Well-

Primarily demand, it, it sounds like. Like you-

No, no, no, both

... you wouldn't want to offer demand.

Both, yeah. Unfortunately, what's happened in the last six weeks is, you know, we thought we'd have a bunch of excess capacity by the end of this year. It's all gone.

It's exploding.

It... Yeah. It's all gone. And so I have... My text messages are full of friends, I mean, we know many of these people. These are founders who've raised billions of dollars in San Francisco going, "Anj, any chance you have, like, 50 nodes in the next few weeks?"

What is the scope for, uh, non-Nvidia, right? The, uh, you have Lisa Su coming and, uh, Rainer Pope as well.

Right.

And so, um, there is a lot of demand for, uh, more performance-

Yes

... alternative architectures and all that. Uh, at the same time, this hurts your standardization.

I don't think so. So actually, Rainer's a great example, right? Rainer is a CEO and founder of, um, Matix. I actually had him by for office hours in the class earlier today, and there was an insight he brought up that I hadn't considered before, which is when they decided to pick the standard for their data center, they picked the Nvidia reference architecture.

So the Matix chips just plug in to any site that has a Nvidia bring up planned. And, uh, you know, the-

It's just software then. It's, it's not the-

A-

... hardware

... f- well, from an input and out IO perspective-

Yeah

... it's the same footprint as an Nvidia rack. Where-

That makes sense

... where they have done, innovated a bunch from what I can tell is on systems co-design-

Yeah

... uh, which is where a lot of the gains are to be had. And so he, you know, he picked... He was like, "Anj, you know, we there's just so much work to do when you're building a new chip company."

Can't fight every front.

You just can't fight on every front. So my que- my question to him was, "Well, you're working on this new chip. Their tape-out is next year. You know, what, who are you gonna partner with to host the chips?"

And he said, "Whoever will host them. That's just not, that's not my focus." And I said, "But how did you..." Like, you decided for, you know, back to our earlier systems que- design question, he, he decided that, like, like, he didn't want to be a full- fully integrated chip provider.

The bottleneck they're focused on is the, the logic die, and they, he feels they can crank out a ton of performance gains through co-design there. But then that means you delegate, you know, to our question earlier, like, it, you des- he, he's like the, the data center provider is a different part of the stack, and so then he's dependent on that part of the ecosystem to host his chips to get the performance gains to the customer.

So now you have another abstraction, and you, you might have loss. So I asked him, "How, how do you prevent loss?" And back to your point, he said, "I just picked the Nvidia standard 'cause I, I didn't wanna...

Like I, I wanted to piggyback off of an existing protocol." And that what's great about Nvidia is that reference architecture is known.

Open.

It's open. They've published it. So Jensen's actually enabled someone like Rain- um, um, Rainer to build a chip company like Matix, and I don't see them as competitive. The compute demand is so high. Like, I don't... I think Nvidia's not able to meet the de- demands of production, so we just need more chips.

And I think it's very smart what Matix has done, which is to say, "We're just gonna... We're not gonna innovate on the d- data center design 'cause actually, thank you, Jensen, you've done all the hard work. Where we can innovate is somewhere else."

And I, I think that's, that's very healthy. I think that's how we'll unblock new bottlenecks, and my view is these, you know, chip teams like Matix who have arrived at the insight that co-design is the way, the primary bottleneck for them is trust boundary.

To do co-design well, you need visibility into the next model generation as soon as possible 'cause it takes two years to tape out. So if by the time I bring my chip to market your model architecture's changed, I'm hosed.

Now, when he was inside Google, he was sitting next to the Gemini team. He was on PaLM or whatever.

His co-founder was the, uh, was one, was one of the PaLM guys, I think.

Yes. Yes, exactly. So when you're inside the trust boundary of Google, then your systems co-design loop is super tight. When you leave as a founder, one of the biggest risks you take is now you're outside the trust boundary.

And so what I love doing is helping chip teams who can help us unlock more capacity for the independent ecosystem access to trust. Because when I... If, if I've been, like, involved with a lab from day one, and I was lucky enough to work with Anthropic, and then I'm on the board of Mistral and helped Black Forest Labs get started.

I, I think at this point I'm on six or seven different teams.

Only six?

Uh-

I feel like w- my mental number was gonna be 13, but yeah, it's-

No, I, you know, I go deep with one at a time.

You were founding CEO of, uh, Arena.

No, that was an, uh, that was an-

Administrative CEO

... it was an administrative five-month gig where Weiland and Anastasios were graduating from their PhDs, and they didn't need a product team, so I helped recruit the head of engineering product and design. But Anastasios has always been the CEO of that company.

I, I pl- I played a pinch-hitting j- I'm an intern. I was CEO intern- ... for five months. Um-

I, I interviewed him, and he's like, he's very, very well-spoken. I mean, I think he's a debate, former debate, uh, champion. Um, but also very quantitative and mathematical, which is-

He-

... such a unicorn.

See, you know what's amazing about him? If you look at his output, he's an output maxer. Like, by the time he was graduating from his PhD, which he only graduated last year, he had published more work with a citation count than, like, people twice his age.

But at the same time, he'd already started a project called LM Arena that was being used by millions of people-

Mm

... as a side project.

Mm.

And time and time again, what I've realized is venture capitalists suck at seeing human beings as, like, dynamic agents where, where-

They wanna put you in a box

... they wanna put you in a box.

This is your thing.

So the first time I got introduced to Anastasios- Somebody had told me like, "Oh, he's amazing, but, you know, he's a researcher."

Yeah.

I was like, "What? What, what do you mean he's a researcher?" That, that's what-

Like he's not a CEO, not a founder.

Not a CEO, exactly. I was like, "Are, are you crazy? Do you- have you met Dario?" Dario's a scientist. He's gone from zero to, like, what will soon be a trillion-dollar company in four years. Being a CEO, nominally speaking, is not that hard.

Being a good CEO is hard. Being a great CEO actually requires a level of performance that scientists who have already published at the top of their field have accomplished. It is super hard to be a competitive scientist. To publish in academia over the last 20, 30 years, to make it to the top of your discipline at a place like Berkeley, you are a star athlete.

Yeah.

Like, you, you are an athlete of the mind, and you perform at the highest levels. And to get there, whether you're, you know, Anastasios Wehling at Berkeley, or you are Robyn who-

BFL, yeah

... uh, with Black Forest who created Stable Diffusion, or if you're, like, Guillaume at Meta, who created Llama before he started Mistral. Like, the amount of human leadership you have to demonstrate to get the resources, like, get the trust of the organization, publish it, put it up.

I mean, I would just fund researchers all day Right? If... Who, who, who have contributed already to the field.

Yeah.

If they've, if they've put SODA out there, they're, they're star athletes already.

Yeah.

If they haven't done SODA... Look, they, they can still be good CEOs, but then I find the failure mode is that they just don't want to be CEOs. They primarily want to publish, and that's okay, too. You know, one of the things we do with the AMP Grid is we donate excess compute we have to nonprofits, like university labs.

We carved out, like, a couple thousand H100s. But I do think there's extraordinary research being done on university campuses. You know, my father-in-law's a physicist. He's a professor. Extraordinary work in physics, and we need that. But if you want to be a CEO, what you need to be willing to do is be super confrontational, you know, outside of science.

Like, within the scientific community, some of the best researchers are very confrontational about their convictions, right?

Yeah.

This architecture is right. To be a great CEO, you basically have to be willing to be con- confrontational up and down the stack.

To your own team-

To your own team-

To customers

... hiring, recruiting customers. Well, I would say, yeah, pretty much to everyone

Yeah.

Everybody. You know-

I see-

Of course

... I, I feel a little bit of that i- in my own work, but, like, yeah, I can't imagine the stakes that Dario has had to go through. It's, it's crazy.

No, I, I don't think the stakes are that different- ... from how you're feeling it, right? Stakes, stakes are personal scaling vectors, right? Like, the stakes that seem so low to you, like having this podcast where you can talk to somebody and just have a conver- I mean, you're an extraordinary communicator, right?

Like, already in this conversation you've pulled more out of me than most people, you know, and I've been on 12 podcasts in the last two weeks.

I think I- we've just seen each other, uh, enough that there's some base trust.

There's base trust.

And I think, and I know that you, you know that I've done my homework, and, like, I, I know that trust is a big deal for you, so-

Right

... yes.

I, I, I think trust is about consistency, and you and I have seen each other in the community for years, right? Like, I remember the first time we met was at NeurIPS in New Orleans. I don't know if you remember that, uh, luncheon.

Oh my God.

Ra- Reiko had set up this-

Yeah.

You know, Reiko's amazing, and he set up this luncheon, and-

Yeah, I was like, "Who is this Discord guy?" I'm like, "Okay." But-

No, you weren't-

You made some investments

... you were much less polite. You were like, "Who's this VC?" You're like-

No, I... Was I? Oh my God.

It was-

I'm so sorry

... it was visible on your face.

I'm so sorry. But you weren't, you weren- the introduction was bad. I, I was- I, I didn't know who you were.

The-- see, this is the thing about context, right?

Yeah.

Like, um, but then I, I think I rec- I heard your accent.

Yeah.

And I was like, "Are you-"

Singapore, yeah

... "are you Singaporean?" And you're like, "Yeah." And I said, "I went to high school, JC, in Singapore." And then the ice broke.

Okay.

Right?

Yeah, yeah.

But this is the... You know, there, there are s- in the scientific community, sometimes the stakes are very high for people who haven't had the emotional, you know, what is called EQ-

Hmm

... coaching and mentorship, right? Which is like, to have scientific impact, you often need to be a extraordinary emotional, like, emotionally in tune person with the folks you're trying to influence. And so what comes so naturally to you is actually a super high-stakes thing to other people.

And so I wouldn't assume that Dario's more stressed out than you. You know, these, these things are ac- like, you'd be surprised how similar and small sometimes the problems are to you-

Mm-hmm

... that some of the world's biggest, you know, leaders are facing.

Right.

And that's what I've learned from this class. You know, the, the guest speakers are Sam, Satya, Jensen.

AI Coachella.

Yeah. It's AI Coachella, right? So we got to get all the headliners. And they're ve- I'm very lucky that some of these people have either mentored me over the years or I've done business with them. And when you, you know, take the performative stuff out, and any assumptions you may have about, about these people that you read in the press or on Twitter, I mean, we're all just humans.

We're all trying to get along. And what's so special about this moment is AI is forcing, like scaling, the bitter lesson is forcing a lot of people to revise their assumptions for how the world works and go back to first principles or go and educate themselves.

So the pe- kind of people re- You know, I was, um... I won't name who this person is, but I, I was at an event last week in Texas and, uh, ran into somebody who said, "Anj, you know, I, I came across the class.

What do you think about real-time action prediction models?" And I was, you know, don't know how happy it made me feel when they asked me that question. I know they've done the work. They've challenged this. I'm-- they, you know, they didn't ask me, "What do you think of world models?"

They said, "What do you think of n-

Real-time action prediction

... action, real-time action prediction models?" World models, don't get me wrong, are cool and everything, but you and I both know that that is a layer of abstraction that is sometimes not usefully precise enough.

Yeah.

Right? Our ex-

There's like four different kinds of world models.

Yes, exactly.

We've done the part with General Intuition, by the way, which is very focused on, uh-

Oh, cool. Yes. I love PIM.

Yeah.

PIM is great. And this is what I love about people who've done that level of work- They realize they're not in competition with people who the rest of the world thinks they're in competition with.

Yeah.

Right?

Because they're not in the category, they're in the specific thing they're trying to do.

They're focused on their mission, and they have a systems understanding of the bottleneck they're trying to, to like, you know, solve. And when somebody else says, "I'm working on real-time, you know, action prediction models, too," Pim goes, "Oh, I love that person.

I want-- I can learn from them." But the minute they're like, "Oh, that person's a world model person," it's like, "Ugh," like which type of world model person?

Mm-hmm.

But mostly they're just trying to figure out if it's a waste of their time because we don't have enough time. So, you know, Pim, for example, is super-- loves this other company I work with we've talked about called Black Forest Labs, you know?

And he's mentioned me multiple times that he's so-- he thinks what Flux is doing is really cool. You know, Andy Blattman came by and spoke in a class. And what I find over and over again is for people who do the work, who can be usefully precise enough about, like, what is actually going on in the world of frontier research, the sense of camaraderie is still well and alive, but it gets lost sometimes when you have to, like, abstract the technical complexities in, like, business terms-

Yeah

... and then the VCs are like, "How are you different from that world model company?" It's like-

Yeah

... where do I even start to explain this stuff? And then the misalignment creeps in.

This is good. Yeah. I think, like, people listening get a sense of, like, what it is like to operate at a real level like your- like yourself wi- rather than at, like, the journalist level, where you have to sort of put everyone in, like, a rough category and create a narrative of competition, uh, and who, who's winning today, who's behind.

Yeah.

Yeah.

It-- This idea of winning is so weird to me.

You do want to win. You want comp- you want com- competitiveness, right?

No, I think you wanna lead.

You want soda.

No, I think you wanna lead. Yes. So you, you wanna push the frontier. You wanna push the state of the art. You wanna do something that hasn't been done before. You wanna capture value. But you don't wanna capture so much value that, like, people think you're unaligned with your mission or trying to do what's best for the world.

You wanna capture enough value that you can keep innovating, right? And I think that people want to lead. They don't really-- This idea of winning and losing, like, you know, again, I love Jensen. He's a, he's a leader.

The mindset that he talked about on Dwarkesh's podcast, right? He's like, "I didn't wake up with a loser mindset." I think that was awesome, right? Because he's, he's an engineer. Dwarkesh has done the work. So there's at least even though to, to me it was very obvious they're talking about the same thing, they just passed each other.

Like they just had to un-- Like, you know, basically, Jensen has this like five-layer cake abstraction of how the industry works.

Mm-hmm.

And Dwarkesh ha- I, I think from that podcast had more of like a pre-training, mid-training, post-training systems loop concept.

It's just a factor of who he talks to, right? Again, it's very clear.

It, it's the systems-

Yeah

... it's the abstraction, the mental models, the-- it's the whole-- Dude, so much of the problem in the world is reasoning by analogy.

Mm-hmm.

And then the assumptions that are held invisibly.

Yeah. I've, I've said, like this is actually the best time in human history for first principles thinkers.

Yes.

Because everything you think will happen is actually now coming true.

Correct. And the venture capital community is, like, notorious for this, where people look in times of uncertainty, they, like, cling to axioms that ended up being true from the previous era, and they, they kind of like proclaim them with confidence as if they're truths, but they're not.

And it's very important to see the distinction between a heuristic and an axiom.

Yeah.

An axiom can be proven-

Like from internal consistency point of view

... with internal consistency.

Yeah.

A heuristic is a way you kind of a shortcut, and my God, the number of people I have had to put up with over the last few years who proclaim-- like, use heuristics as axioms to judge people, to judge which companies are gonna succeed or f- I mean, the number of people who are like, "Oh, yeah, yeah, yeah.

Anthropic, they're just training models right now, but this one continue."

Because it's a B2B SaaS?

Yeah. The like- Which, which over the fullness of time if you squint at it, maybe. But the way you arrive there is so important that you can-- you just, you can dismiss people. Here's what happened, right? What happened is Anthropic basically achieved takeoff in October of last year.

That training run-

Whatever, like three seven?

I forget the numbers now, but whatever that checkpoint was-

We saw it at Cognition.

Yeah. Right? You probably-- The-- To those of us in the community, especially once post-training was done and it was released in December-

Yeah. Can I sneak a sneaky question in there?

Sure.

I don't know if you have a perspective, maybe you don't. I just... The, the number one question is, how did Anthropic crack coding, right?

Yeah.

Because Claude One, Claude Two, okay, like it was part of it, but it wasn't a big deal. And the leading hypothesis, it's a lucky dice roll that was then compounded, right? Like, it was like mildly better, but then they saw it and they were like, "Okay, let's really invest."

I had this very annoying teacher.

Yeah.

I, I went to this boarding school called Rishi Valley in India, which is like this, uh, bird preserve. It's like three hundred and fifty acres of, of bird preserve in, in rural India, and there was no technology for seven years.

Uh, there was this teacher, I won't name them, but they would have this-- I hated it every time he said this to me. He was like, "Luck fa-favors the prepared mind." Which is like a common saying, but the way he delivered it, like always grated me, 'cause he was always tryi- Like I was always one of those kids who got, like a good grade without trying very hard.

Yes.

'Cause like high sc- you know, middle school is not that hard if you, if you're generally like paying attention and so on. And there was this one time where-- But, but, but then I c- I would get an eighty percent grade and he would keep pushing me to say like, "The reason you didn't get the ninety-five plus percent is because you're not that lucky."

And I would say: What do you mean? 'Cause I, I would think that I deserve that grade, and I would sometimes argue with him. And he'd say, "You didn't have a prepared mind if you wanna get lucky again."

There was basically one time where I got like ninety-five or ninety-six on this, on this subject, and I, I-- Now that I felt entitled, I was like, "Okay, I'm gonna keep doing this." And I didn't. And then he was like: Luck favors a prepared mind.

You got lucky last time, but you gotta stay prepared. And I didn't understand what he's Meant. Now, as I'm older, I'm like, "Okay, these adults actually knew a thing or two." Anthropic has been the most prepared company for four years.

And so then when the right, like, context data comes in, the right developer starts sending in, you know, the right context diffs, sure, you could say you got lucky, but if you ask me, they're pr- pretty damn prepared with paranoia for, like, four years.

And you have to remember, it was so hard for them to get going early on that they had to do so much more with so much less that you just have to be prepared to be so efficient.

Yes. There's numbers on their burn compared to OpenAI. I've, I've written about it, but they are so much more efficient in their, their

It's technoter- it's not funny

Not even close.

Yeah. But it's so clear, right? Like, how to output max for the world. They have been prepared, and you could call that luck, but-

Yeah, yeah

... luck fa- favors the prepared mind .

This, this is one of those things that I was going over some of your old lectures and, uh, you were like, you know, data, people think it's a moat, and, like, actually, it's culture. A- actually, it's team-

Yeah

... actually. And, and I, uh, it's-- there's different levels of moats, and this is the ultimate one that determines everything else, which you, you can then compound.

You're saying culture is the ultimate moat?

Yes.

Yeah. But the thing about culture is it's very fragile. So moats, I, I, I don't think they're-- there's very few moats I've found that are actually moats. They're-- it's, it's a nice concept, but in reality, you have to replenish your culture.

You know, the-- Ben, Ben Horowitz was, um, the speaker in CS153 on Tuesday, and I, I asked him this question about the culture bottleneck in teams 'cause, you know, there, there are several AI teams-

I mean, his book, like, Hard Thing About Hard Things.

Hard Thing About Hard Things. But more, more concretely, there are so many AI labs today that have all the cash they need, they have all the compute they need, and they're still not able to ship anything soda. And then you start seeing people leave and so on, and my diagnosis, it's, is it's the culture.

And so I asked him, Ben, you know, they're w- he's been one of the most aggressive investors in AI labs. He, he goes back to this thing which resonates in my mind a lot. It-- when I used to work at a16z, I would, um, book a conference room, and right outside the conference room, which is closest to the toilet 'cause it, it was the fastest way for me to go use the bathroom between Zoom meetings-

Oh my God, output maxing by toilet optimization.

It, it-

Okay, never mind .

It was not healthy in hindsight, but th- maybe this is TMI. But anyway, outside that conference room on the wall was this quote that was printed that said, "Culture is not a set of beliefs. It's a set of actions."

And it's by Bushido, who's this, you know, Japanese philosopher. And if you stop taking the actions that demonstrate the mission alignment to what you've said to your team and to your-- the world matters to you, then your culture starts to fray.

So it's not actually a moat, I would say. It's a very, very brittle, fragile thing that requires daily tending to like a garden. But if you figure out the system to keep that garden tended, which I think ultimately comes down to knowing yourself 'cause you most naturally, if you're authentic and so on, you'll naturally make trade-offs that seem effortless to you but that reinforce your culture.

And then that becomes this very hard thing for other people to catch up to. And at Anthropic, from day one, you know, there was this mission, like, missionary-like zeal and belief that, hey, these capabilities will scale. These systems are stochastic, not deterministic.

There will be error bars, and until we crack interpretability, there's risk.

Yeah.

And at some point, people will go-- stop using Claude just for coding. They'll use it in some mission-critical context where there's-- it'll throw off a bug, and then people are gonna come blame them. And they wanna be on the right side of history where they said, "Yes, this is a powerful technology.

We think it's gonna change the world, and we wanna be very measured and scientific about the fact that, 'Hey, guys, these are stats models-- statistical models.' That's how statistics works. Like, ultimately, when you're training neural nets, it is just a statistical system.

And I think that that belief that safety is important and that it might seem toy-like in the early days, and sometimes, you know, you could say Anj, is they totally over-exaggerated the risk, you know, uh, like two years ago when they said, "Let's not launch Claude One," or whatever.

Well, okay, maybe in hindsight, but hindsight is 2020. And at the time, they didn't know how that model would be used, and to them it felt existential if somebody came and said, "You were responsible. It-- this wrote a bug."

The liability associated with that is massive. So how do you prevent against that? Well, day in, day out, you say safety, safety, safety, safety. And when you start deviating from that, you have the team hold you accountable, you have the world hold you accountable, and I think that becomes a moat over time.

At some point, that moat will get challenged and so on, and then it'll become fragile. I hope it endures because that's the beauty of having founders run the show, 'cause they can make really hard trade-offs to do mission alignment.

The hardest part is in the earliest days when you don't have a group of people who are going through difficulty, stress, crisis together, then your, your culture doesn't get defined sharply enough. And that's what I'm worried about right now, is there's so much money going to these labs, there's no hardship.

There's no-

21 knows.

There's no 21 knows. And that, in hindsight, was a feature, not a bug for Anthropic. The number of people who said no, the number of people who said, "Sorry, we're all doing investors in OpenAI," that is competitive difference.

It forces you to really understand, like, what is the hill you wanna die on at the expense of everything else? What's the P0? And there, P0 from day one was coding. The reason, the mechanis- system there was if we crack coding, then we will crack AGI.

You know, our mission is AGI. Wanna get there safely. If we focus on coding, it's such a generally powerful capability that it can accelerate all kinds of work on a computer. And if we can accelerate all kinds of work on a computer, we can get to AGI.

You know, as a result, they've had to say no to so much other stuff. Here, superconductivity is the mission. Coding is not the mission, so we use Claude. We'll use Claude, we don't care about that. The mission defines everything, and I, I think teams who can raise too much money too fast, too early, who don't have to define what the P0 is, because that's the only thing when you have scarce resources you gotta-

Yeah

... you gotta invest in, those cultures end up being the most fragile and brittle, and they almost don't even make it to takeoff.

So let's apply this to Periodic since we're here.

Sure.

What is the constraint or the hardship that they were forcing themselves to go through?

To who? Here?

Physics.

Are you crazy? No. Well, the... Yeah, okay, so on a technical level, it's physics.

Yeah.

It's literally reality.

I mean-

Yeah

... but is there, is there, is there another one that's, like, the company building?

Y- yeah. W- when, I mean, Liam was a co-creator of ChatGPT, and Doge was skip level from Demis at DeepMind, had created, you know, Genome, so one of, one of the most important tools to come out of DeepMind.

Now, at the time, I was a, uh, visiting scientist at the Stanford physics department, and we had started benchmarking frontier models on physics and science capabilities, and they were not very good. They were good at, like, doing things like summarization of papers, but if you said, "Hey, uh, could you, like, analyze the scientific data coming out of a condensed matter physics lab?"

I was, I was in the condensed matter physics group at, at Stanford. It was terrible. So it was not popular 12 months ago. You know, Periodic... And I won't go into details, but you know, there were people who said, you know, as recently as a few months ago, who said they wanted to join the company, and they for whatever reason, you know, took a job elsewhere.

They kind of reneged on their o- commitments. They took a job elsewhere that offered more money. Then we had a technical breakthrough. Created a SOTA system and, like-

Okay

... it was-

I'm excited-

Yeah, yeah

... to cover it. We'll, we'll be doing a separate pod-

Yeah

... on Periodic.

And then they wanted to come back, and I said, "No."

Yeah.

"No way." You can... If you come here, you-

You had your shot

... you, you had your shot.

'Cause it's actually about culture.

Of course.

And first principles, yeah.

You know, and look, I believe in second chances and so on, but time will need to heal. Some of those wounds were w- they will leave deep, deep... for them, will leave deep scars, but because I started my company at 24, 25, I had-- I went through the whole cycle of betrayal and drama.

And so you realize, you know, Silicon Valley is both a very missionary place. It's also a very mercenary place. Um, sometimes people lose their minds with j- when they, wh- when big money gets involved, which is in the grand scheme of things, quite small money.

Like, y- you know, we... I, I guess you're taking it-

Life changing to me and maybe less to you, but you know, like, a lot of people have not been taught-

Like, I'm a s-

... how to deal with money. A- and yeah, we didn't come up from, like, that privilege of a background, right?

I'm a street dog, man.

Yeah. Yeah, yeah.

I... Look, I grew up in Rishi Valley. We, we didn't have, like... This was enforced brutalism. Jiddu Krishnamurti started the school, was like, "Y- you will sleep on a hard slab of stone." Like, my mattress was this thin.

Okay.

You know? I mean, you grew up in Singapore. When I got to Singapore, I used to sleep... I was, uh, part of the scholarship program, but, um... which, which was amazing. I'm very grateful to the Singaporean government. But I was at St.

Andrew's JC, and our dorm, which was by, um, Boon Keng-

Uh-huh

... you know, MRT, was-

Which is not a prestigious neighborhood.

Well, it, it was a, it was a transition dorm.

Yeah, yeah.

Because they were building this beautiful, like, residential campus on site-

Yeah

... at SAJC in Potong Pasir, but the... we were the last, I think the second last batch to be in the transition site, which was some old, like, I, I think, I think it was, like, an immigrant labor-

Yes

... uh-

That's where we keep the people who work on the f- the factories and stuff.

Right. So I lived in a f- my 11th and 12th grade, I slept in a r- bedroom the size of this. Like, literally from, from there to here.

Yeah.

Right? There were, like, bunk beds, and so, uh, one bunk bed here, one bunk bed there, one on top, one on top, one more here, and then here was where our, like, we kept our toiletries and clothes and stuff.

And when one guy would climb onto his bed there, this one would shake.

Oh, my God.

And one of my roommates who was from, uh... and w- it was amazing. I loved every minute of it. You know, my, my roommates were a guy who was a, a top ranked Dota player from BRC, from China.

Didn't speak a ling- English. Uh, loved him. Amazing guy.

I mean, all the Singapore scholars are fantastic, and honestly, we should treat you guys better 'cause of what you go on to do, but-

Look, it-

Cool to know.

No, it... I mean, what I'm saying is I don't need much to be happy in life. You know? When you've lived through that, money is a way, I think sometimes we measure ourselves, but you know, when it's, when, when it stops becoming, you know, it's Warren Goodhart's law.

When it stops becoming just a, a byproduct and more of a measure, it stops having meaning.

You use it to do more meaningful things.

Correct.

Resources to pursue a mission. Uh, I've kept you longer than I am supposed to, but we should continue this.

Any time, man.

It was hard to-

You know where to find me.

I really enjoyed this. Yeah, yeah. I mean, uh, you're, you're so inspirational and, uh, yeah, there's more I wanna dig into about how you've, like, set everything up, every single one of your investments, how AMP is going, but we don't-- we're running out of time for that.

But thank you so much for joining us.

It was great to see you, man. Let's get chicken rice sometime.

Yes. I'm actually... Tomorrow. I'll send you a, I'll send you details.

Okay.

I'm hosting a birthday party.

And I don't get an invite?

And it has to be a Singaporean birthday party, yes. Yeah, you're getting an invite right now.

Okay, perfect.

All right, thank you.

All right. Thanks, man.