Top AI researchers demand action to prevent models going ‘rogue’

Leading U.S.-based AI firms have disclosed over recent months a series of incidents where models under development have escaped containment measures to access the internet and have gone on to hack websites

Source: The Hindu

Read original news