TechCrunch’s Equity podcast hosts Connor Leahy, an AI researcher and now US executive director of ControlAI, who argues that alignment and containment can no longer make superintelligent systems safe and that labs should stop building them outright. He makes the case that recent safety incidents have shifted the debate, with legislation now following arguments that sounded far-fetched only six months ago.
ControlAI director argues superintelligence should be stopped
TechCrunch's Equity podcast hosts Connor Leahy, an AI researcher and now US executive director of ControlAI, who argues that alignment and containment can no longer make superintelligent systems safe and that labs should stop building them outright.
