> It's patently insane to demand that humans alter their behavior to accommodate the foibles of mere machines

Did you fully read the original thing? No demands were being made, or I didn't read it that way. It was simply a suggestion for a better way of interacting with AI, as it stated in the conclusion:

"I am hoping that with these three simple laws, we can encourage our fellow humans to pause and reflect on how they interact with modern AI systems"

Sure, (many/most) humans are gonna do what they're gonna do. They'll happily break laws. They'll break boundaries you set. Do we just scrap all of that?

Worthwhile checking yourself here. It feels like you've set up a straw man.

> There is no finite set of rules that can constrain AI systems to make them "safe". I don't have a proof, but I believe that "AI safety" is inherently impossible, a contradiction of terms. Nothing that can be described as "intelligent" can be made to be safe.

If we want to talk about "disagree with this framing", to me this is the prime example. I'm struggling to read it as anything other than defeatist or pedantic (about the term "safe"). When we talk about something keeping us "safe", we're typically not saying something will be "perfectly safe". I think it's rare to have a safety system that keeps you 100% safe. Seat belts are a safety device that can increase your safety in cars, but they can still fail. Traffic laws are established (largely) to create safety in the movement of people and all the modes of transportation, but accidents still happen.

I'm not an expert on this topic, so I won't make any claims about these three laws and their impact on safety, but largely I would say they're encouraging people to think critically. I'd say that's a good suggestion for interacting with just about anything. And to be clear, "critical thinking" to me means being skeptical (/ actively questioning), while remaining objective and curious.

Not a real argument or anything, but I'm reminded of the episode of The Office where Michael Scott listens to the GPS without thinking and drives into the lake. The second law in the article would have prevented that :)