So on that front, while I didn’t guess this, it does sound like a good investment.
Furthermore, while I am skeptical about the “DC in space” thing in the short term, and Musk’s timelines are, generously, something like P99.9 outcomes, it seems plausible that in 10 years it might make sense. And at that point Google will want to put their TPUs in orbit, too.
The earth is limited in the amount of thermal energy it can radiate.
Rather than co-locating media servers into ISPs you can stick them into space so they're just closer.
Space in NYC isn't cheap. Space in space is free.
... what?
Say what you want about starship etc etc, but space in space is far from free and will be for a long time.
> Data centers in space?
Everything else you said draws from some regular consumer experiences. That's not what I am talking about here.
This may be a narrative violation on HN, but the quality of Google's management is exceptional.
In-house is Google's moat. The big piece right now is LLMs where in-house TPUs allow cheaper internal costs. If Cloud customers want to rent Nvidia GPUs that's fine since they're paying. But Search and etc get an edge on competitors by not.
The bigger problem with in-house is the failure to out-house it. Kubernetes shouldn't been new. Things should be done like AWS where the internal APIs are the external APIs.
I don't think anybody at Google has sent that email, yet. Maybe Sergey and Sundar don't use Google products, anymore. Maybe they have a special Google just for billionaires that doesn't suck.