What if AI is asbestos?

Published 2025-05-29

Human civilization has a decent track record of putting dangerous technologies back in the bag. Asbestos is still used only in very specific and regulated cases. X-ray machines are no longer present in shoe stores to check shoe fit. Nuclear non-proliferation treaties have so far forestalled armegeddon.

Regulation is geographic, but all-in-all, the regulation is still effective for people who live in regulated places. If there’s a developing country somewhere that’s making shoes out of asbestos or something, that sucks for them, but it doesn’t have a huge impact on me.

If it were somehow proven that AI would have a significant negative net impact on society, would it be possible to ban or control it in any meaningful way—one that makes a difference on a global scale? Can any digital entity truly be banned?

I’m not asserting that AI is or will be a net negative—that’s beyond my pay grade—but what if?