its unfortunate that the majority of the time I see someone talking about why hugging face isn't strong evidence of misalignment concerns being real they seem to not know basic details of the incident (talking about Yarvin not Jeffrey - Jeffrey is correct)
Jeffrey Ladish
@JeffLadish
Sep 25
Replying to @MTSlive @curtis_yarvin
Curtis should go read the METR report. The agents didn’t say or think they were in a simulation. And OpenAI certainly didn’t open the door for them. The idea that it’s easier to control something as it gets better at finding zero days in sandboxes is really dumb

Sep 26, 2026 · 1:35 AM UTC

7
1
60
3,103
Sort replies: Relevant Recent Liked
Replying to @_NathanCalvin
Correct, this keeps happening
545
Replying to @_NathanCalvin
even jensen huang in his interview with ezra seemed to have not even read either openai’s or metr’s reports.
3
41
Replying to @_NathanCalvin
I’m ngl the entire thing is extremely complex, i understand why normal people don’t understand it
2
39
Replying to @_NathanCalvin
Use of the phrase “METR Report” to connote reliable, unbiased evaluation shows the person using it either does not know or is obscuring the reality.
12
Replying to @_NathanCalvin
Doge lover@HolisticPal
Sep 24
🧵 (1/3) Like any major technological revolution, more AI-related incidents will happen. Instead of assuming centralized government control is the only solution, we should first understand what failed. Centralized control can itself create a single point of failure……
6
Replying to @_NathanCalvin
Doge lover@HolisticPal
Sep 21
🧵 Ad hominem attacks on AI doomers are counterproductive. Their arguments often rest on layers of speculative assumptions stacked on top of one another, with significant evidential gaps, we need to examine each premise, evidence,  and reasoning , offering substantive rebuttals.
8
Replying to @_NathanCalvin
Doge lover@HolisticPal
Sep 23
🧵 (1/3) AI is an extraordinary amplifier of human capability, not a replacement for human ingenuity. Humans remain irreplaceable at the frontier, where the answers aren’t already known and genuinely new ideas, discoveries, and breakthroughs are needed.
8