Gavin Baker - Nvidia v. Google, Scaling Laws, and the Economics of AI

“Application SaaS companies are making the exact same mistake that brick and mortar retailers did with e-commerce. Brick and mortar retailers looked at Amazon and said, it's losing money. E-commerce is going to be a low margin business. How can it ever be more efficient? Our customers pay to transport themselves to the store and then they pay to transport the goods home. Amazon's vision was, we're just going to go down a street and drop off a package at every house.”— Gavin Baker

Strategy & Decision Making · Business & Entrepreneurship · Economics & Markets

DUR_ENDURING

SaaS repeating retailer mistake

“Any function where there's a right or wrong answer or a right or wrong outcome, you can apply reinforcement learning and make the AI really good at that. Does the model balance? They'll be really good at making models. Do all the books globally reconcile? They'll be really good at accounting, double entry bookkeeping. It has to balance. Support or sale. Did you make the sale or not? That's just like AlphaGo. Did you win or you lose? Did the guy convert or not?”— Gavin Baker

Technology & Engineering · Business & Entrepreneurship

DUR_ENDURING

Verifiable functions are automatable

“Having watts as a constraint is really good for the most advanced compute players because if watts are the constraint, the price you pay for compute is irrelevant. The TCO of your compute is absolutely irrelevant because if you could get 3x or 4x or 5x more tokens per watt, that is literally 3 or 4x or 5x more revenue. If you're going to build an advanced data center costs $50 billion, a data center with the ASIC maybe costs $35 billion.”— Gavin Baker

Economics & Markets · Strategy & Decision Making

DUR_ENDURING

Constraint makes efficiency trump cost

“With software, anything you can specify, you can automate. With AI, anything you can verify, you can automate. It's such an important concept. And I think important distinction. Verified is such an important concept in AI. One of Karpathy's great things was with software, anything you can specify, you can automate. With AI, anything you can verify, you can automate.”— Andrej Karpathy

Technology & Engineering · Philosophy & Reasoning

DUR_ENDURING

Verify not specify

“What Google has been doing as the low-cost producer is they have been sucking the economic oxygen out of the AI ecosystem, which is an extremely rational strategy for them and for anyone who's the low-cost producer. Let's make life really hard for our competitors. All of that calculus changes once Google is no longer the low-cost producer, which I think will be the case.”— Gavin Baker

Strategy & Decision Making · Economics & Markets

DUR_ENDURING

Low-cost producer can suck oxygen

“Data centers should be in space. What are the fundamental inputs to running a data center? There are power and there are cooling. In space, you can keep a satellite in the sun 24 hours a day, and the sun is 30% more intense. This results in 6 times more irradiance in outer space than on planet Earth. So you're getting a lot of solar energy. Because you're in the sun 24 hours a day, you don't need a battery. The lowest cost energy available in our solar system is solar energy in space.”— Gavin Baker

Technology & Engineering · Philosophy & Reasoning

DUR_ENDURING

Space solves power and cooling

“Had reasoning not come along, there would have been no AI progress from mid-2024 through essentially Gemini 3. There would've been none. Everything would've stalled. Can you imagine what that would've meant to the markets? For sure, we would've lived in a very different environment. So reasoning kind of bridged this 18-month gap.”— Gavin Baker

Technology & Engineering · Strategy & Decision Making · Economics & Markets

DUR_CONTEXTUAL

Reasoning bridged 18-month hardware gap

“No one on planet Earth knows how or why scaling laws for pre-training work. It's actually not a law, it's an empirical observation. It's an empirical observation that we've measured extremely precisely and has held for a long time. But our understanding of scaling laws for pre-training is kind of like the ancient British people's understanding of the sun, or the ancient Egyptians' understanding of the sun.”— Gavin Baker

Technology & Engineering · Philosophy & Reasoning · History & Geopolitics

DUR_ENDURING

Measurement without understanding

“I was a ski bum at Alta in college. I was a housekeeper. I've cleaned a lot of toilets. It was shocking to me how people treated me. You'd be cleaning somebody's room and they'd be in it and they'd be reading the same book as you. And you'd say, oh, that's a great book. You know, I'm about where you are. And they look at you like you're a space alien. Like you speak. And then they get even more shocked. You read? So it had a big impact on how I've just treated everyone since then.”— Gavin Baker

Psychology & Behavior · Leadership & Management

DUR_ENDURING

Humility from being treated poorly

“It takes at least 3 generations to make a good chip. The TPU V1, I mean, it was an achievement in that they made it. It was really not till TPU v3 or v4 that the TPU started to become even vaguely competitive. Is that just a classic, like, learning by doing thing? 100%. The first ASIC team at any semiconductor company is the Amazon ASIC team. Trainium and Infantry One, maybe they're a little better than the TPU v1, but only a little. Trainium Two, you get a little better.”— Gavin Baker

Technology & Engineering · Operations & Execution

DUR_ENDURING

Three generations to make good chip

“Blackwell was by far the most complex product transition we've ever gone through in technology. Going from Hopper to Blackwell, first you go from air-cooled to liquid-cooled. The rack goes from weighing round numbers, 1,000 pounds to 3,000 pounds, goes from round numbers, 30 kilowatts, which is 30 American homes to 130 kilowatts, which is 130 American homes.”— Gavin Baker

Technology & Engineering · Operations & Execution

DUR_CONTEXTUAL

Complex transitions need infrastructure overhaul

“Reasoning fundamentally changed the industry dynamics of frontier labs. With reasoning, that flywheel has started to spin. If a lot of people are asking a similar question, they're consistently either liking or not liking the answer, then you can use that, that has a verifiable reward, that's a good outcome. And then you can feed those good answers back into the model. This is important fact number one. Meta threw a lot of money at it and they failed. Microsoft also failed.”— Gavin Baker

Technology & Engineering · Strategy & Decision Making

DUR_CONTEXTUAL

Reasoning enables flywheel

“You have to use it yourself. I'm amazed at how many famous and august investors are reaching really definitive conclusions about AI based on the free tier. The free tier is like you're dealing with a 10-year-old and you're making conclusions about the 10-year-old's capabilities as an adult. You could just pay. You do need to pay for the highest tier, whether it's Gemini Ultra, Super Grok, whatever it is, you have to pay the $200 per month tiers. Those are like a fully fledged 30, 35-year-old.”— Gavin Baker

Technology & Engineering · Strategy & Decision Making · Psychology & Behavior

DUR_ENDURING

Free tier inadequate for evaluation

“AI happens on X. There have been some really memorable moments. Like there was a giant fight between the PyTorch team at Meta and the JAX team at Google on X. The leaders of each lab had to step in publicly. The companies are all commenting on each other's posts. The research papers come out. If on planet Earth there's 500 to 1,000 people who really, really understand this and are at the cutting edge of it, and a good number of them live in China, I think you have to follow those people closely.”— Gavin Baker

Strategy & Decision Making · Technology & Engineering

DUR_ENDURING

Follow the 500-1000 true experts

“Have you ever seen public opinion change so fast in the United States on any issue as nuclear power? Just happened like that. And why did that happen right when AI needed it to happen? Now we're running up on boundaries of power on Earth. All of a sudden, data centers in space. It's just a little strange to me that whenever there is something, a bottleneck that might slow it down, everything accelerates. Rubin is going to be such an easy seamless transition relative to Blackwell.”— Gavin Baker

History & Geopolitics · Culture & Society · Technology & Engineering

DUR_ENDURING

Bottlenecks generate solutions

“Shortages are always followed by gluts in capital cycles. What if in this case the shortages compute. There will eventually be a glut. AI is fundamentally different than software, in that every time you use AI takes compute in a way that traditional software just did not. I think we went into great detail on maybe a prior podcast about how just inventory dynamics made these inventory cycles inevitable in semis.”— Gavin Baker

Economics & Markets · Technology & Engineering

DUR_ENDURING

AI uses compute per query

“Investing is a game of skill and chance, kind of like poker. There's obviously chance in investing. If you're an investor in a company and a meteor hits their headquarters, that's bad luck, but you own that outcome. So there is chance that is irreducible, but there's skill too. The way you got an edge in this greatest game of skill and chance imaginable was you had the most thorough knowledge possible of history, and you intersected that with the most accurate understanding of current events in...”— Gavin Baker

Strategy & Decision Making · Economics & Markets

DUR_ENDURING

History plus current events equals edge

“I am a competitive person and I've actually never been good at anything else. I got picked last for every sports team. I love to ski. I've literally spent a small fortune private skiing lessons. I'm not that good of a skier. I like to play ping pong. All my friends could beat me. I tried to get really good at chess. Never beat one of them. I've never been good at anything. I thought I would be good at this.”— Gavin Baker

Psychology & Behavior

DUR_ENDURING

Only good at one thing