alt Hacker NewsThe problem is recursive self improvement creating a very difficult gap and the fact that power, compute has a lag from when you invest and when data centers come up.
You also can't just spin up a research team out of nowhere.
The problem is recursive self improvement creating a very difficult gap and the fact that power, compute has a lag from when you invest and when data centers come up.
You also can't just spin up a research team out of nowhere.
These things are true, however:
1. The labs in the US and China don't seem to have any problem selling (or even giving) access to these models right now.
2. If some kind of take-off happens which makes that not true, my bet is all bets are off on what that outcome even looks like. What would the economic paradigm even be under a superintelligent AGI? Do you think "it" is going to listen when Trump says "you can't work with Europe"?
There's a whole bunch of grey in between the two, for example only having access to second rate models, but I'm not sure that particularly matters if the strategy is "second mover."