once again it needs to be said:
THERE IS NO EVIDENCE THAT AI COULD DESTROY THE WORLD. YOU MADE IT UP IN YOUR HEAD
This isn't EA vs e/acc. It's Longtermists vs. Utilitarians.
Longtermists are against coin flip bets to double or destroy the world. Risk neutral Benthamites like SBF are for them.
If you think the p(doom) from superintelligence is non-trivial but the upside makes it worth accelerating anyway, you're closer to the EA, Singerite ideal of a sociopathic utility maxer.
In this sense, SBF and Altman are more the same than different. They're the true EAs marinating in the ego death of abstract utility calculations, while the Longtermists are the true human loving civilizationalists. Civilizationalists understand the intergenerational value of growth and development, but for that very reason reject the myopic, high time preference logic behind games of Russian Roulette. To be a literal accelerationist, in contrast, is to be the 17 year old in their dad's car who, in a vain attempt to impress their friends, ends their life wrapped around a telephone poll.