Was he a bit teary saying those words? youtube.com/watch?v=zuZ2za…

Aligning AI is a dead end.

Want to get up to speed on AI? My top recommendations are: - "On the Dangers of Stochastic Parrots..." - EU AI Act - "AI: Unexplainable, Unpredictable, Uncontrollable" - Gary Marcus's newsletter - PauseAI website - Yudkowski's episode on Bankless - Wiki page on TESCREAL

Bindu Reddy "Safe superintelligence" in the pipeline right next to "drinkable sand" and "tasty manure". x.com/alexalarga/sta…

Add your name to our statement calling on governments to implement what polling shows the public already want – a global pause on frontier AI development.

AI automation is NOT just an economic issue! Labor doesn't just give you money, it also gives you power. When the world doesn't rely on people power anymore, the risk of oppression goes up. Right now, popular uprisings can and do regularly overthrow oppressive governments.

If you are a Bayesian, you do not update your priors (typically). You update your beliefs. Or you update from your priors. Or acceptably update upon your priors. Just saying that if you want to sound fancy, get your grammar right.

Recklessly endangering life (individually or as part of a group one freely chooses to associate with) is inconsistent with being a good person. Discounts can be made for ignorance, lack of understanding, etc. These don't apply because they hire people who should know better.

No. You simply have not seen radical yet. May we never see radical. The pause position is the most moderate, unbloody position out there.

genuine question, who even uses this?

Laura Tingle's comments are deeply offensive and grossly irresponsible. The ABC has a statutory obligation to disseminate news and information impartially and accurately. That one of the ABC's most senior journalists says this murderous terrorist attack "has nothing to do

This is a post about AI

How do you regulate a threat that evolves so fast? #PauseAI We need time to work out what’s safe and what isn’t, and if we don’t Pause frontier AI development, we won’t have time. It will take long enough to deal with what’s already been released.

With its gain-of-function research, Anthropic is the Wuhan Institute of AI.