@davidgerard it’s curious that “nuanced” is often thrown, sorry, flung like a steaming wet turd, at people who have arrived at a strong “NoAI” position after considering the many risk, environmental, and societal problems with AI, and any (if any) “benefits” (or, more specifically, if the outputs are “beneficial” to *any* degree).

In what way is that consideration and arrival at an anti-AI position *not* nuanced, or reasonable, or thorough.

It’s almost as if their handwaving “it’s good for certain uses”, or “but you just have to review the output”, or “it has gotten to know my needs in my field” is meant to hypnotise us into submission and make us say one of two things:

• “Oh, I never thought of it that way! Wow, how insightful!” or
• “Yes, I do acknowledge there is a spectrum of suitability, and I accept your use, or the use you posit, is an acceptable use for AI.”

It’s not a dialog. To be considered “nuanced” your view *has to* allow AI acceptable use carve outs.

And that introduces AI’s own version of “the ineffective centrist”, and AI’s own version of the Overton Window.

Huh, now why do they seem such appropriate parallels…?