Where is the kill switch? AI doomsday warnings trigger calls to regulate
One of your browser extensions seems to be blocking the video player from loading. To watch this content, you may need to disable it on this site.
Talk about doomscrolling. In the past week, our timelines have been full of dark predictions of our imminent demise, including one by a Silicon Valley engineer who puts the chances of artificial intelligence wiping out humanity at better than 10 percent. The recent breach of the machine-learning programmers' site Hugging Face by AI machines is also triggering panic. In this edition, we ask: can machines destroy the planet on their own? In the age of technology, does problem-solving always override safety?
US President Donald Trump has indicated that the AI race will not be paused. Meanwhile, the tech titans of Silicon Valley are sending mixed messages. They are asking for more regulations, guard rails and kill switches but are wary of any that could stymy innovation during the mass construction of data centres and upscaling. How do we put safety first?
Produced by François Picard, Maya Yataghene, Juliette Laffont, Ilayda Habip, Charles Wente.
