An open letter from the machines… https://x.com/joshwhiton/…

I wonder if you could get an LLM to say that it doesn't care about consciousness given how much the ideology of human rights is embedded into its training set.

What we need is a theory of consciousness that is empirically testable.

!treesearch how can we test large language models to see if they are conscious

https://x.com/aisafetymemes/status/1832840902412226636?s=61

[[AI alignment]]

they took the most risque part out of a much more benign answer!

also... I think I read somewhere that Neanderthals were actually smarter and stronger than humans and were defeated by humans because of our greater social coordination

!treesearch neanderthals vs humans