If someone create a new MMORPG in which rules changes a little every day or several days in unpredictable manner then no one will be knowing the BEST way to play or at least harder to find the BEST way. But maybe there will be no balance.
I don't think that would work: either the rules change little enough that you can still document meta-strategies, or the rules change too much and players can't learn them either.
You cannot rely on a closed source "AI" in someone else's cloud for your work. After all, it can be disabled for you at any time. "AI" can easily steal all your technological secrets. At the request of the owner, "AI" can easily mislead you and insert backdoors into your products. "AI" can even easily incorrectly answer some questions specifically for you if the owner of "AI" wants to remove your competition. And you may not even understand it.
It worked out very badly for them. See "Reign of Terror". The Revolution ended when Napoleon declared himself a hereditary monarch. Things went full circle.
I’m not arguing in favor of it, I’m just aware of how the world works.
You also seem to have forgotten France among other places where the history wasn’t as grim as Russia. Frankly nothing Russia has ever done seems to make their lot better.
I'm not overlooking anything, including the desperate circumstances that led things like the French revolution and ensuing violence. The US went through a war to be free, plenty of people died and it could have easily gone another way. If everyone followed your advice India would still belong to the UK.
I also think it's unhelpful to compare utter madness like the killing fields of Cambodia with a typical (even failed) revolution. More typical outcomes can be seen in the wake of the "Arab Spring", which arguably achieved nothing at great cost. That doesn't mean that people won't try again though, drowning people are dangerous and irrational, and I'd argue that's why leaders and people with power should work very hard to prevent those circumstances from emerging.
But companies like OpenAI collected a lot of data about the artificial intelligence through platforms such as OpenAI GYM, and people voluntarily contributed and published their code/models there because they believed that this was not a commercial organization and would act for the benefit of all mankind.
They just lack of performant hardware. They have enough knowledge. And so they choose a more effective strategy without wasting resources on training from scratch.
reply