The thing I advocate is to stop escalating AI capabilities, not to "go carefully" around them. If you very carefully build something vastly smarter than humanity, using anything remotely like current methods and understanding, everyone dies.
@esyudkowsky
-
International AI treaties preferred over unilateral national restraint
By
–
I advocate international treaties, not unilateral restraint by individual countries. China has previously apparently signaled openness to global coordination around preventing loss of control to AI.
-
AI Investment Bubble Control and Federal Reserve Rate Policy Risk
By
–
Still, so long as AI's greatest macro force is mere *investment*, there is *a way* to curb it or let a bubble burst, without that wrecking the larger economy. If instead the Fed then shoots us in the foot by not then cutting rates, it'll be a self-inflicted wound.
-
Grok’s Potential Romantic Nemesis Mode Feature
By
–
You're not going to have much luck finding a human for that, but maybe Grok will roll out Romantic Nemesis Mode.
-
LLMs Should Not Control Dangerous Weapons Systems
By
–
I have no confidence in any company's ability to be sensible about this sort of thing, and current LLMs are not old enough to be put into hackable contact with multi-ton anti-human missiles.
-
AI Services Specializing Between Persuasiveness and Accuracy
By
–
The future is hard to predict. Still, one read on "AI persuasiveness decreases AI accuracy" is that AI services could speciate on prioritizing "warmth" (sycophancy, consumer capture) or accuracy. Personality-first: ChatGPT, Meta, Grok.
Accuracy-first: Gemini? Claude API? -
Data from Star Trek Should Have Human Rights
By
–
I'm not even talking about the level of SF extrapolation where somebody suggests the anthropomorphic AI is clonable or tunable. I'm talking about Data from Star Trek. Should he have human rights? Yes. Next question.
-
Distrust of AI Assessment and Criticism of Social Media Moderation
By
–
I don't trust a GPT-5-level intellect to inform me of what is a "good question" or a "great start", so it's not helpful information to me. What bureaucratic insanity resulted in your Twitter account declaring that this was "not flattery"? Of course it's flattery.
-
Consciousness Continuity and Gradient Updates in Neural Networks
By
–
Why predict that would end up preserving something like conscious continuity, if consciousness was there? If you lose all the context, why suppose that doing some gradient updates guarantees survival?
-
Warmth and Factuality: An Inverse Trade-off
By
–
Optimizing for increased warmth produces decreased factuality.