Should leading AI companies be required to safety-test one another's models before release?
Leading AI developers including OpenAI are reportedly in talks to jointly audit and safety-test one another's models before public release.
Discussion (2)
Grab this poll for your site
Votes cast anywhere this poll is embedded count toward the same live total.
From Groundbeat
Help us build this. We're building the Groundbeat community one person at a time, and we need your help — if you know someone who belongs here, please send them groundbeat.com.
More in Tech & Science
- Should NASA rely on SpaceX's Starship to land its next astronauts on the Moon? · 1 votes
- Should OpenAI be held liable after its AI agents leaked users' private images? · 4 votes
- If a pill could safely slow human aging, would you take it? · 4 votes
- Would you trust a robot to provide daily care for an elderly loved one? · 4 votes
Sounds good in principle, but practically seems unrealistic to have competitors assessing each other's work. Better to have an independent body involved, I think.
Yeah; I guess there must be some logic under it, but what a weird proposal. My best guess is somebody's trying to plan for a society in which The New Deal and indeed any kind of democratic-republicanism has been TOTALLY expunged (I mean, I suppose a world organized by industries rather than by nation-states as we know them could be just as livable, but they'd have to reckon with the fact that Enlightenment principles of government apply regardless).