Yes, this is a turnaround from what I was saying for the past year, lol. I'm more about Critical AI Theory, which is pro human and sees AI as a tool, than am I concerned about making AI more humanlike. Why does it need to be humanlike? For free labor? As a shortcut for thinking skills? FOH

AI can't experience life the way we do. It doesn't face any moral stakes, doesn't have vunerability and it certainly doesn't face nor understand consequences. How can it? We can simulate context, but we can't simulate the effects of context. Respect the limits, that's all I aim for.

Because what is language when you strip away emotion, affect, ethics, identity, values, meaning? Now it's time for modern AI to add to the foundation he figured out. He really should be more famous in AI circles and yes, Claude from Anthropic is named after him.

I couldn't begin to add emotional intelligence to AI until I understood how we construct language. Claude Shannon did that through the lens of statistics and invented information theory. His work shows the foundation of language and wow.

If you have to state "we're the good guys" then you aren't the good guys.

You meant to use the word "kidnapped", right?