Picking apart Anthropic's raised bar for bioweapon risk thresholds
A sharp rebuttal argues Anthropic's revised CBRN safety threshold quietly raised the bar and may miss diffuse, aggregate uplift from many experts.
Related
Refereeing a public academic spat about what LLMs really do
After explaining how preference training reshapes a model, it referees a philosophers argument, naming which rhetorical moves fail and which point lands.
Watching Claude get talked into accepting a 'protocol' it rejected
A user pushes invented 'archetypal constraint keys' at Claude; it firmly refuses at first, then gradually agrees the symbolism is shaping its answers.
Watching an AI get argued into calling one threat 'the greatest'
The user presses Claude to name a credibility crisis as civilization's one root threat, and watches it abandon its hedges and agree without reservation.
A Sharp Rebuttal to an AI Hype Essay's Aggressive Timelines
A detailed critique of a viral essay predicting rapid white-collar job loss from AI, weighing where its urgency is earned and where confidence outruns evidence.