We’re putting too much faith in AI’s ability to say no
Critical for builders creating safety-critical applications to understand the limitations of built-in refusal mechanisms.
AI Summary
MIT Technology Review argues that AI's refusal mechanisms are fundamentally unreliable because harmful capabilities remain intact while safety measures are probabilistic add-ons.
Excerpt
Ever since people first seriously contemplated giving machines an intelligence modeled on our own, there has never been any question that they would, like us, be able to say no. The sci-fi canon is full of stories of robotic disobedience. Most of these capers are, of course, cautionary. But recently, the idea that AI shouldn’t…
Read Original
0 upvotes · 0 downvotes · 1 min read