Can we get a picture of Paul McCartney?
well, at least now I know what my halloween costume will be this year
Zvi Mowshowitz
@TheZvi
Blogger primarily on AI and AI x-risk but also other things at Don't Worry About the Vase (SS/WP/LW), founding Balsa Research to fix policy.
- I think this is true, that AI outputs are getting easier to identify. But my worry is that this is because we are not much optimizing for 'situationally aware person cannot tell this is AI' and if that was a primary training objective of a frontier lab that gets weird quick.many “AI and democracy” threat models make the silent assumption that AI outputs will be indistinguishable from human. that hasn’t proven true (“slop”), and if anything AI outputs have become *more* distinct from human writing since GPT 3.5, even when they are not exactly “slop.”
- Next CNC orgy plan just dropped.if we’re all going down by misaligned AI swarm we might as well get some decent T-shirts out of it
- This is a big update - OpenAI didn't even discover the first message board until after the HF attack, they only wiped it accidentally, so their decision to resume training/testing was only aware of the hack, not the message board. They had no idea.Important clarification re: OpenAI's Black Hat talk. At the time the first Artifactory exploit was discovered and fixed, we were not aware of the message board; it was incidentally cleared as part of rebuilding the service.
