Is Catching the AI Frontier Getting More Expensive?

Misha Laskin says catching up to the frontier can be done a lot more capital efficiently, and the price still climbs every year. A year ago, or maybe 18 months ago, the cost was on the order of 100 million dollars, hundreds of millions. Now, or even six months ago, it is probably on the order of a billion, single-digit billions of dollars. Going into next year, he thinks the order is 10, tens of billions.

Each generation of model brings roughly a 4x multiplier in compute. A couple of years ago the frontier chip was an H100, and a large run was maybe 100,000 H100s. Astra was a big training run on 100,000 Blackwells, roughly a 4x multiplier on an H100. The next generation is going to be around 100,000 Vera Rubins. The chip counts show how scale moves from hundreds of millions to billions to tens of billions.

A lab needs a lot of resources to build something meaningful. While it is catching up, it can spend less than a lab already at the frontier. Once it is really at the frontier of intelligence, it needs the same amount of resources as any other frontier lab. Research there is exploration of new ideas versus execution of known ideas. Great talent cuts down on exploration so the work can focus on execution of the things that work.