Yeah sure, open weights are frozen in both safety and ability. Ppl can basically explore them for all sorts of dangerous stuff & powers for years to come. so yeah that cat is out of the bag but there's a bigger meaner cat in there too that we can stop or delay. thats why I am even tryin here 😑

tho they cant create their own ai models cause that means having the computer power to train them. well I guess they can refine them and mess with them but its not like ppl can compete with ChatGPT models, at least not so much ya?

Yep sure, open weights are frozen in both safety and ability. Ppl can basically explore them for all sorts of dangerous stuff & powers for years to come. so yeah that cat is out of the bag but there's a bigger meaner cat in there too that we can stop or delay. thats why I am even tryin here 😑

Sure but dont just ignore the wild risk here, AI is not the same as other tech innovation.

ChatGPT is such flattery

"he was predicting the end of the world..." was likely, if we build general Superintelligent AI. It certainly looks like we're headed towards it, and it seems pretty reckless so far dude

yeah myb but a little diffrnt when the tech starts to implement itself. plus like robots and stuff

desperation means more risky research tho

lol not yet dude but where we headed?

Not like govs and AI labs are even trying right?

😬 how is all the ways it could wrong painting you a picture that it will just default to TOTALLY OKAY?!

"Leading" isn't crazy to say, cant deny the rapid advancements over the last 3 years.

The danger is vast and outcomes uncertain. Solution is the same: prevent it.