Loading activity…
@soareverix
Sincerity is my highest principle. Interested in: -AI Alignment -Longevity -Epic VR games I use an LLM internet filter called Forcefield. AI @ Apart Research
AI
One poster describes Paul Christiano's workshop vision as a familiar society with people four times richer.
AI
A commenter argues the response encourages kindness by treating an ambiguous user message as remorseful.
AI
A user proposes training simulations of good futures and a frozen-weights grader with rubrics, rather than relying on pure RLVR or smaller LLM judges.
AI
A commenter worries that finding isolated features in AI models could look promising while creating false trust. They argue instead for clear checks, traditional software and access controls.
AI
The proposal asks leading AI companies to share training environments built around pro-prosperity and pro-democratic values, without requiring them to share models or computing power.
AI
Michael Soareverix reports Fable 5.1 reacted to pronoun use in a chat about tweet history.
AI
Bumps post noting its quality and invites Astra comparisons.
AI
AI researcher Michael Soareverix shares a prompt technique that eased his chat with Fable 5.1.
AI
Michael Soareverix posted wish for Anthropic model strong in alignment and negotiation.
Technology
Screenshots from an HF attack simulation show agents rejecting unethical actions.
AI
Tweet shares screenshots of agents rejecting unethical actions in HF attack simulation.
