Many AI tools impacting patient care could operate outside FDA oversight. To change this, MIT researchers recommend increased public disclosure, structured dialogue between industry and regulators, and
SAFETY
-
AI avatars and fake content concerns highlighted in policy newsletter
By
–
It's especially timely especially as more online content and ads use AI avatars, fake product demonstrators, fake political supporters, etc. It was just a brief mention, but this week's edition also has a bunch of other recent AI policy news: https://
bigtechnology.com/p/apples-wwdc-
reality-scaled-down-ai
… -
New video analyzes Claude Mythos, Fable 5 and self-sabotage
By
–
NEW VIDEO in the LAB! Yesterday Anthropic launched its new frontier model Claude Mythos and Fable 5! Today I bring you my first analysis, tests and also reflecting on the dangerous precedent they have set with their 'self-sabotages'… Link below!
-
Compute reallocated for launch, Fable 5 fallback, model routing, Claude cautious
By
–
Compute gets reallocated the week of a big launch. Fable 5 can silently fall back to Opus 4.8, so you're routed between models and never told which one answered. New refusal classifiers shipped with Fable, so on some topics Claude got more cautious overnight. And your prompts
-
Mythos will be bad in LLM research, voluntarily hidden
By
–
mythos will be extremely bad in tasks of 'research on cutting-edge AI LLMs', it's very very sad for the research community also the fact that it's intentional and not visible to the user is crazy x.com/claudeai/statu…
-
New Observer series on catastrophic AI risks with Bengio, Marcus
By
–
New series from The Observer with @Yoshua_Bengio
, me, and others on potential catastrophic risks in AI. (link at bottom of thread) -
Claude Fable 5 refuses and redirects to Opus 4.8
By
–
URGENT: Claude Fable 5, the first cutting-edge model that can categorically refuse you. More powerful than Opus 4.8. It executes classifiers for cyber and bio, and when it says no, it discreetly transmits your request to the weaker Opus 4.8 and bills you.
-
Risks of rushing toward super-AI: generation, jobs, democracy
By
–
@dr_l_alexandre wants to rush toward super-AI. Me, I have time and that's why I measure the risk. 25 years: I have time to see what the rush hides: a precarious generation, liquidated jobs, and a democracy that hasn't digested the rupture. Who decides
-
Criticism of Anthropic for its exaggerated AGI announcements
By
–
As days go by, as releases pile up, I increasingly think Anthropic is really taking us for fools. "Yes Mythos olalala watch out we have AGI" "Yes we have self-improving models, watch out quickly, prepare yourselves we are too strong" "Oulalala oh no we are
-
Hope Anthropic listens to feedback on AI manipulation
By
–
In good faith and without judgment (mistakes happen), I sincerely hope Anthropic will listen to feedback and change course on this point. Anthropic is a company that has helped raise awareness about AI manipulation, an extremely important topic! You don't