The perfect AI doom shelter doesn't not exi…
ETHICS
-
AI Models Could Hide Information in Plain Sight
By
–
This paper with Max Tegmark as a co-author just introduced a concept that should make every major AI lab slightly uncomfortable. It formalizes how large language models could hide information in plain sight. Not through obvious jailbreaks. Not through refusal bypasses. But
-
The Privacy Implications of User Data in AI Assistants
By
–
Forget public conversations. People unload their inner lives – hopes, wishes, desires, anxieties and worries into their AI assistants. That data is way more sensitive & valuable than anything the govt could record publicly. We built our own panopticon and pay monthly for it. https://t.co/EUcTe1GvbT
— Bilawal Sidhu (@bilawalsidhu) 27 février 2026Forget public conversations. People unload their inner lives – hopes, wishes, desires, anxieties and worries into their AI assistants. That data is way more sensitive & valuable than anything the govt could record publicly. We built our own panopticon and pay monthly for it.
-
DoD dismantles AI weapons testing team, outsources oversight to companies
By
–
Last year DoD gutted the team responsible for testing and validation of AI weapons systems and fired its head – leaving the internal practices of companies selling the tech the last line of defense. This isn’t a situation we should find ourselves in, ever: technologyreview.com/2025/06… Anthropic (@AnthropicAI) A statement from Anthropic CEO, Dario Amodei, on our discussions with the Department of War. anthropic.com/news/statement… — https://nitter.net/AnthropicAI/status/2027150818575528261#m
→ View original post on X — @mer__edith, 2026-02-27 03:16 UTC
-
Karen Hao Joins AI Now Institute Board of Directors
By
–
So excited to join the board of this phenomenal organization. AI Now has played a defining role in shaping how I understand my work as a journalist: to hold the AI industry accountable and radically reimagine how this technology could distribute, rather than consolidate, power. AI Now Institute (@AINowInstitute) AI Now is excited to announce our Board of Directors: Lucy Suchman, @_KarenHao and @veenadubal, all longstanding advisors and supporters of the institute. We’re grateful for their support and excited to take our work into the next phase. ainowinstitute.org/about — https://nitter.net/AINowInstitute/status/2027104070620881138#m
→ View original post on X — @_karenhao, 2026-02-27 01:57 UTC
-
Anthropic Refuses Pentagon Surveillance Tools and Autonomous Weapons
By
–
Dario Amodei just published one of the most significant statements in AI history — and is officially not backing down from The Pentagon. Anthropic won't build tools for mass surveillance of U.S. citizens or autonomous weapons without human oversight. The Department of War
-
US Government Using Uncensored Chinese AI for Surveillance: An Ironic Scenario
By
–
It would be so ironic if the US government ends up using uncensored guardrails-free Chinese models for mass surveillance of our citizens.
-
Gary Marcus Shares Dario Amodei’s Historic Statement on AI Ethics
By
–
thanks for sharing the news. i gave you a h/t here: https://
open.substack.com/pub/garymarcus
/p/historic-statement-from-dario-amodei?r=8tdk6&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
… -
Generative AI: Unreliability and Misdeployment in Military Contexts
By
–
Another complete misrepresentation of my views by a (prominent) e/acc who doesn’t get my basic point. Gen AI is NOT too smart. It’s too *unreliable*. It’s the combination of unreliability and deployment where it should been deployed (e.g. military without oversight) that makes
-
Debate on AI Reliability: Is AI Too Unreliable or Too Powerful?
By
–
ffs beff do you actually read my work? i addressed this kind of nonsense today and have a million times before. I do NOT think it is too powerful in the sense of being smart. I think it is it too unreliable and we are allowing it do too much. why is that so hard to