> I'd argue that being bought by a government and controlled within the ethical confines of that government would be a great way for OpenAI to quickly fall behind its peers who aren't subject to such regulations (which is the entire point of this post in the first place).
Behind in what? You write as if it is a univariate concern. I haven't read the post but why couldn't Norway force OpenAI to be more alignment-conscious? Then it would be ahead in safety, especially if competitors don't pay attention to that.
“Behind” = less capable, less used, less resourced. Safety and alignment don’t necessarily correlate with market dominance. If they did, no safety regulations would be needed.
We have to stop pretending that the market automatically arrives at socially optimal solutions. We don't say car safety regulations need to be weakened to compete with China, but when it comes to AI, safety magically goes out the window.