logoalt Hacker News

drivingmenutsyesterday at 3:02 PM1 replyview on HN

In an ideal world, a product that can be harmful is tested privately until there is a reasonable amount of safety in using that product. With AI, it seems like that protocol has been completely discarded in favor of smoke-testing it on the public and damn the consequences.

Of course, investors are throwing so much money at AI and AI is, in turn, buying legislators and heads of government, who are bound and determined to shield them from liability, so …

We are so screwed.


Replies

aejtaetjyesterday at 3:15 PM

Every underdog will skimp on safety, especially those in other jurisdictions.