This story originally appeared on Mother Jones and is part of the Climate Desk collaboration.
\n\nWe’ve been hearing a great deal recently about the specter of AI-driven human extinction. Maybe our misplaced trust in AI intelligence will set off World War III. Or AI agents will unleash widespread cyberattacks and take over the internet. Given too much autonomy, the machines will run amok. They may seek to deceive their creators—wait, that’s already happened. Suppose they conclude, like Agent Smith in The Matrix, that “human beings … are a plague, and we are the cure.” Or perhaps the chatbots help bad humans unleash a literal plague—a biological weapon.
\n\nThat last one isn’t merely theoretical. A “misuse” report Anthropic made public earlier this month detailed five cases in which foreign scientists tried to evade Claude’s guardrails on “gain of function” inquiries. That’s a euphemism for when scientists tinker with a pathogen’s genes to enhance it in some way, making it more infectious, deadlier, or better able to spread among mammals, for example.
\n\nAnthropic denied those users the answers they sought and cut off their accounts, even while noting that their motives may not have been sinister. Such information also can be used…
Original source: https://www.wired.com/