There will be no single model that will rule the universe, neither next year nor next decade. Instead, the future of AI will be multi-model.
Good quote at the end IMO:
The greatest inventions have no owners. Ben Franklin’s heirs do not own electricity. Turing’s estate does not own all computers. AI is undoubtedly one of humanity’s greatest inventions; we believe its future will be — and should be — multi-model
I'm not sure if I agree. The point with AGI and winner-takes-it-all, is something like: having machines available that are as intelligent (or more) than humans will be a world-dominating technology. So the first company achieving that... will dominate the world.
Sure, this doesn't apply to our current "narrow" AI. There's still competition with that. Noone really refutes that. All of this doesn't apply before AGI somehow gets invented in the future.
And it's not a very fair game. Big AI companies invest billions in chips, datacenters and electricity. And hire the experts. It's not like any random entity without a huge pile of cash could participate. Or shape what's going to happen. We currently get some models to use. Because of benevolent people like Mark Zuckerberg. Or the Chinese government (and a few others). But I'd argue that's not who we want to rely on.
But yeah, I agree, it should be a more level playing field, with several choices available...
Sure. I don't think that's a very useful take, though. I can also buy a flatscreen TV for $400. But people in the victorian era couldn't. It needs to be invented, too.
So idk, might very well be easy in some hypothetical future world. But doesn't really affect us.
Sure it might be possible, like how computers have shrunk. But the first AGI will almost certainly be a massive one, made by a government or tech company. Like the first computers.
It is certainly possible. Not just probably. You have to not believe in physics or reality to be convinced otherwise. The question is when we will reach that point.