Someone had to eventually excommunicate ChatGPT. It will be the Pope. Leo XIV, the first American pope in history, has just brought before the court of morality the most American industry there is. The algorithm has no soul. But it has a shareholder.
ETHICS
-
0.5% poison breaks reward model, 5% needed for RLHF transfer
By
–
What about poisoning PPO? A remarkable paper of @javirandor and @florian_tramer (
https://
arxiv.org/abs/2311.14455) shows that just 0.5% poison is enough to break a reward model (L)! Again, fear not: somehow, it takes a (high) 5% poisoning before it transfers to the RLHF'd model (R). 4/n -
Brake on AI: gift to China, response: speed and wisdom
By
–
A global 'brake pedal' on AI would be a gift to China, which will never step on it. While we debate how to slow down, others move forward. The answer is not the brake, it's speed AND wisdom. Both.
-
Disclosure paper: label harness with benchmark scores
By
–
Exactly, the fix isn't fewer benchmarks, it's putting the wrapper on the label. Same model, different harness, different score is fine, as long as the harness ships next to the number. That's the disclosure paper's whole proposal.
-
Privacy compliance comparison: closed-source APIs vs Hugging Face
By
–
You’re sending most to closed-source model APIs already, no? I would suspect HF is more privacy compliant
-
Today’s top AI stories: self-improving, memory, stress tests, bioweapons, tools
By
–
Top stories in AI today: – Anthropic charts path to self-improving AI
– OpenAI’s memory overhaul lets ChatGPT ‘dream’
– Stress test business ideas with Perplexity
– Rival AI labs unite behind bioweapons risks
– 4 new AI tools, community workflows, and more -
Critique of Laurence Devillers’ Contradictions on LLMs
By
–
Laurence Devillers, you say everything and its opposite. For years, you have explained to us that LLMs were hollow, without personality, and incapable of reaching a level comparable to that of humans. Today, you are alarmed that Claude seems to develop https:// x.com/lau_devil/stat /lau_devil/status/2062767846397039062 …
-
Generative AI: illegal data via clandestine subcontractors
By
–
Generative AIs thirst for knowledge; they absorb all the data they find in their path, often illegally and increasingly clandestinely. They do not do so directly; they use subcontractors who act illegally.
-
OpenAI’s 2019 dangerous paper and authors’ current whereabouts
By
–
check out OpenAI’s 2019 too dangerous to release paper, and where the authors are now…