The “Kill Switch” Problem
Why one of the most-used phrases in AI safety policy carries several different meanings, how polysemous language is now being voted into law, and the realities of what a shutdown can and cannot do.
#policy
3 posts tagged "policy"
Optimizing to Human Rights Over Innovation
A critique of the ‘Innovation First’ approach to AI, arguing that human rights should take precedence over innovation when the two are in tension.
Lessons Learned in Edinburgh
Early reflections on the programme — why ethics needs alliances, the power of virtue ethics, and growing market demand for AI ethicists.