Antonio Gracias argues Google will win in AI because it has its own custom chips (TPUs), a strong focus on ROIC, and a monopoly ad business to fund infrastructure — making the $75B CapEx justifiable on a return-on-capital basis.
Google will also win because they have their TensorFlow chips. They make some of their own chips. They do focus on ROIC, and they have a great monopoly to kind of fund it all.” ⚑
Friedberg views Google's $75B CapEx as a positive signal, arguing Google has historically been the most frugal and well-managed infrastructure investor, with strong ROIC discipline and a credible line of sight to monetize the investment — and the spend demonstrates confidence that search can evolve to chat.
I would view the $75 billion CapEx actually as a very positive signal for the company. I think that it means that they have a really strong line of sight on how they're going to have full utilization and great return on this.” ⚑
Chamath argues Google has the best broad-based AI models and a money machine in ads that directly benefits from AI-driven optimization, but needs better disclosure on how its $75B CapEx is allocated between high-return ad optimization and more speculative pre-training.
Google's models are probably the best of all the models across a broad base of capabilities if you test for them... The other thing that Google has is a money machine that directly benefits from these AI-driven optimizations on ad” ⚑