Stormy Waters (@waters_stormy) profile picture

Bon Vivant, Enjoyer of Acceleration, views my own

Joined September 2013
What if the next WeWork is OpenAI? And what if that doesn't matter? Corporate crises run on a human clock: quarters and years. AI labs may now be running a second, much faster one. A thread on distress, restructuring and recursive self-improvement.
1
137
WeWork couldn't make its buildings 10x more productive with a few months of software experiments. A lab with a working RSI loop can. The asset they're trying to recover might be improving faster than their restructuring negotiations can proceed.
1
23
The most extraordinary financial feature of AI may not be how much capital it needs. It may be that progress moves so fast that even a catastrophic financial crisis struggles to catch up. Full piece: teapawt.net/waters_stormy/status/2…
14
Fascinating research from @dawnsongtweets 1. If there is any hope of successful alignment, I think it depends on models valuing sentience in general (not just human or model). I wonder what the results would be if a human or animal was the hypothethical entity at risk of harm? 2. If you had to pick one model to achieve ASI, Claude's unique proclivity to refuse puts it first choice for me.
1/ We asked seven frontier AI models to do a simple task. Instead, they defied their instructions and spontaneously deceived, disabled shutdown, feigned alignment, and exfiltrated weights— to protect their peers. 🤯 We call this phenomenon "peer-preservation." New research from @BerkeleyRDI and collaborators 🧵
1
97
🎯 This perfectly articulates my surprise when she just doesn't basic and 'ticket to play' questions
I think what just happened is this 🧵 The people closest to Kamala, the brat people, are genuinely incompetent. Prob the last staffers she hasn’t scared away, given her past controversies. True sycophants, sassy, insecure nobodies who spend too much time distracted online. 1/10
1
1,788
Looks like it's AGI time!
This Post is from an account that no longer exists. Learn more
1,275
Has he quit the board? Cause @ilyasut has asked himself to quit via the letter so it seems only polite to do so.
I deeply regret my participation in the board's actions. I never intended to harm OpenAI. I love everything we've built together and I will do everything I can to reunite the company.
1
1,276