I don't know if SA is an Effective Altruist, but it feels a bit like he got caught up in their weird logic of "AI will be an extinction-level threat to humanity - and the only way to address the threat is somehow to build exactly that AI".