Head of Alignment at OpenAI Resigns

X posts aren't readable without being logged into an account.

"Feel the AGI." That line just isn't getting any less creepy.

The important parts from the Twitter thread:

>I joined because I thought OpenAI would be the best place in the world to do this research. However, I have been disagreeing with OpenAI leadership about the company's core priorities for quite some time, until we finally reached a breaking point.

> I believe much more of our bandwidth should be spent getting ready for the next generations of models, on security, monitoring, preparedness, safety, adversarial robustness, (super)alignment, confidentiality, societal impact, and related topics.

> These problems are quite hard to get right, and I am concerned we aren't on a trajectory to get there.

Damn. This is pretty much confirming everyone's fears that Altman is going full steam ahead without as much care for safety anymore.

I'm not sure what to do about it though.

Well hopefully these models begin to plateau due to running out of data to train on, hitting they max size, or because Microsoft stops subsidizing them and OpenAI can't afford to train them anymore.

But yeah if they keep going the www will probably become useless for people. Especially if someone doesn't figure out a way to deal with the bots. Idk how you'd make a human only web though.

Then again maybe a little less time on the web would be good for everyone.

Jan Leike Resigns from OpenAI - https://news.ycombinator.com/item?id=40363273 - May 2024 (381 comments)