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 is losing money. E-commerce is going to be a low margin business. How can it ever be more efficient? Right now, our customers pay to transport themselves to the store and then they pay to transport the goods home. They did not invest in e-commerce. They clearly saw customer demand for it.”— Gavin Baker

Strategy & Decision Making · Business & Entrepreneurship · History & Geopolitics

DUR_ENDURING

Margin obsession causes strategic failure

“If you have a decisive cost advantage and you are Google and you have search and all these other businesses, why not run AI at a negative 30% margin? It is by far the rational decision. Take the economic oxygen out of the environment. You eventually make it hard for your competitors who need funding, unlike you, to raise the capital they need. And then on the other side of that, maybe you have an extremely dominant share position.”— Gavin Baker

Strategy & Decision Making · Economics & Markets

DUR_ENDURING

Negative margins as competitive weapon

“Google has been the lowest cost producer of tokens. AI is the first time in my career as a tech investor that being the low-cost producer ever mattered. Apple is not worth trillions because they are a low-cost producer of phones. Microsoft is not worth trillions because they are a low-cost producer of software. NVIDIA is not worth trillions because they are the low-cost producer of AI accelerators. It has never mattered.”— Gavin Baker

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

DUR_ENDURING

AI uniquely rewards cost leadership

“The very nature of AI is you have to recompute the answer every time. A good AI company might have gross margins of 40%. The crazy thing is, because of those efficiency gains, they are generating cash way earlier than SaaS companies did historically, not because they have high gross margins, but because they have very few human employees. It is tragic to watch all of these companies.”— Gavin Baker

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

DUR_ENDURING

AI margins low but cash-generative

“With software, anything you can specify, you can automate. With AI, anything you can verify, you can automate. Verified is such an important concept in AI. That is Karpathy's great insight.”— Andrej Karpathy

Technology & Engineering · Philosophy & Reasoning

DUR_ENDURING

Verification enables AI automation

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

Technology & Engineering · Strategy & Decision Making

DUR_ENDURING

Verification criterion for AI tasks

“Scaling laws for pre-training, no one on planet Earth knows how or why they work. It is actually not a law, it is an empirical observation, and it is an empirical observation that we have measured extremely precisely and has held for a long time. 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 still enables progress

“With reasoning, that flywheel has started to spin. If a lot of people are asking a similar question, they are consistently either liking or not liking the answer, you can use that, that has a verifiable reward, that is a good outcome. And then you can feed those good answers back into the model. And we are very early at this flywheel spinning.”— Gavin Baker

Strategy & Decision Making · Technology & Engineering · Business & Entrepreneurship

DUR_ENDURING

Reasoning enables data flywheel

“Google made more conservative design decisions. Google does mostly the front end for the TPU, and then Broadcom does the back end and manages Taiwan Semi. The front end is like the architect of a house. They design the house. The back end is the person who builds the house. And then managing Taiwan Semi is like stamping out that house. For doing those two latter parts, Broadcom earns a 50 to 55% gross margin. If in 2027, TPU is $30 billion, Google is paying Broadcom $15 billion.”— Gavin Baker

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

DUR_ENDURING

Scale justifies vertical integration

“Had reasoning not come along, there would have been no AI progress from mid-2024 through essentially Gemini 3. There would have been none. Everything would have stalled. And can you imagine what that would have meant to the markets? Reasoning kind of bridged this 18-month gap. Reasoning kind of saved AI because it let AI make progress without Blackwell or the next generation of TPU, which were necessary for the scaling laws for pre-training to continue.”— Gavin Baker

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

DUR_CONTEXTUAL

Reasoning bridged hardware gap

“Data centers should be in space. What are the fundamental inputs to running a data center? Power and cooling, and then the chips. In space, you can keep a satellite in the sun 24 hours a day, and the sun is 30% more intense. You can have the satellite always catching the light. The sun is 30% more intense, and this results in 6 times more irradiance in outer space than on planet Earth. Because you are in the sun 24 hours a day, you do not need a battery.”— Gavin Baker

Technology & Engineering · Operations & Execution · Economics & Markets

DUR_ENDURING

Space solves power and cooling

“No one builds data centers faster than Elon. Even once you have the Blackwells, it takes 6 to 9 months to get them performing at the level of Hopper. Because Hopper is finely tuned. Everybody knows how to use it. The software is perfect for it. The engineers know all its quirks. Everybody knows how to architect a Hopper data center at this point. When Hopper came out, it took 6 to 12 months for it to really outperform Ampere.”— Gavin Baker

Operations & Execution · Technology & Engineering

DUR_ENDURING

Speed enables learning and debugging

“Blackwell was delayed. Going from Hopper to Blackwell, you go from air-cooled to liquid-cooled. The rack goes from weighing 1,000 pounds to 3,000 pounds, goes from 30 kilowatts, which is 30 American homes to 130 kilowatts, which is 130 American homes. I analogize it to imagine if to get a new iPhone, you had to change all the outlets in your house to 220-volt, put in a Tesla Powerwall, put in a generator, put in solar panels, put in a whole home humidification system and then reinforce the floor...”— Gavin Baker

Technology & Engineering · Operations & Execution

DUR_ENDURING

Product transitions require infrastructure overhauls

“The free tier is like you are dealing with a 10-year-old and you are 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 is 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. It is really hard to extrapolate from an 8 or a 10-year-old to the 35-year-old. And yet a lot of people are doing that.”— Gavin Baker

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

DUR_ENDURING

Free tier misleads about true capability

“Edge AI is the most plausible and scariest bear case. In 3 years, on a bigger and bulkier phone to fit the amount of DRAM necessary, you will be able to probably run a pruned down version of something like Gemini 5 or Grok 4 at 30, 60 tokens per second. And then that is free. This is clearly Apple's strategy. We are going to be a distributor of AI and we are going to make it privacy safe and run on the phone. And if 30, 60 tokens a second at a 115 IQ is good enough, I think that is a bear case.”— Gavin Baker

Strategy & Decision Making · Technology & Engineering

DUR_CONTEXTUAL

Edge AI threatens cloud compute demand

“In the data center, the racks are over a certain distance connected with fiber optics, and that just means a laser going through a cable. The only thing faster than a laser going through a fiber optic cable is a laser going through absolute vacuum. So if you can link these satellites in space together using lasers, you actually have a faster and more coherent network than in any data center on Earth.”— Gavin Baker

Technology & Engineering

DUR_ENDURING

Vacuum beats fiber for speed

“It is actually really hard to keep a big cluster of GPUs coherent. A lot of these companies were used to running their infrastructure to optimize for cost instead of performance and complexity and keeping the GPUs running at high utilization rate in a big cluster. There are wild variations in how well companies run GPUs. If you have 30% uptime on that cluster and you are competing with somebody who has 90% uptime, you are not even competing.”— Gavin Baker

Operations & Execution · Technology & Engineering

DUR_ENDURING

Uptime creates massive advantage

“Investing is the search for truth. If you find truth first, and you are right about it being a truth, that is how you generate alpha. And it has to be a truth that other people have not yet seen. You are searching for hidden truths. 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 the world to form a differential opinion on wh...”— Gavin Baker

Philosophy & Reasoning · Business & Entrepreneurship · History & Geopolitics

DUR_ENDURING

Alpha from hidden truth

“AI is the first time where every level of the stack that I look at, the most important competitors are public and private. NVIDIA has very important private competitors. Broadcom has important private competitors. Marvell has important private competitors. There is even a wave of innovation in memory, which is really exciting to see because memory is such a gating factor.”— Gavin Baker

Business & Entrepreneurship · Strategy & Decision Making

DUR_CONTEXTUAL

Competition across entire stack

“I have just been fascinated that for the last 2 years, whatever AI needs to keep growing and advancing, it gets. 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 are running up on boundaries of power on Earth. All of a sudden, data centers in space.”— Gavin Baker

Culture & Society · History & Geopolitics

DUR_EPHEMERAL

Mysterious perfect timing