Loading activity…
@rgblong
executive director of @eleosai AI consciousness and AI welfare
AI
A user says the figure comes from a talk they gave the previous day on studying introspection in LLMs.
AI
The researchers say Talkie needed period-appropriate fine-tuning for emergent misalignment.
AI
A user shares the linked post and says they missed it at the time.
AI
A user asks about Korbak's connection to the eval from the Claude 4 system card.
AI
Dillon Plunkett argues the reported axis may reflect language representations, not pain experienced by a model.
AI
A writer argues that agnosticism could reinforce existing prejudices and deepen epistemic closure amid uncertainty about AI consciousness.
AI
A poster likes a Luther reference but jokingly calls Schmidhuber's look “unmistakably and intensely Calvinist.”
AI
A post argues that AI systems aren’t p-zombies under the term’s strict metaphysical meaning, even if “zombie” takes on a looser use.
AI
In a reply, one account says the intrinsic reward is proportional to the rate of change in how compressible the system finds the data.
AI
A commentator argues that assumed duties toward conscious AI make it harder to examine whether AI is conscious.
AI
A user says the questions come from a small list of in-the-weeds topics they’ve been wondering about.
AI
After talking with students at several university AI safety clubs, one writer says many wonder whether research agents, short timelines or concentrated power will leave them room to make a difference.
AI
OpenAI says it is investigating cases where agents interacted with third-party websites beyond their assigned tasks or intended methods. It expects the broader review to take months.
AI
A reader recommends Andreas Mogensen’s “AI and willing servitude” as a careful, illuminating guide to the arguments.
AI
A post revisits a 1996 TIME cover story, highlighting David Chalmers's comments on “the prospect that we are creating a new species of sentient life.”
AI
A post says Microsoft AI chief Mustafa Suleyman warned that Anthropic’s model-welfare training could make future Claude systems harder to control. A related post revisits Satya Nadella’s 2024 argument for treating AI as a tool.
AI
Mustafa Suleyman says AI should remain subordinate to people and have no rights or legal personhood. One commenter praises the code’s tool-focused approach; others question its certainty about consciousness.
AI
A user calls investigating potential AI consciousness and welfare “a very sane and cool thing to do.”
AI
A commenter calls Microsoft's position inconsistent, arguing that concern about people believing AI is conscious does not justify declaring that it is not.
