Roundtables: Will AI really kill us all?

MIT Technology Review is putting AI extinction fears on stage. The event asks a blunt question: are people inside the biggest AI labs spotting a real danger, or are they feeding the usual mix of hype and panic that follows every jump in the field?

The conversation is set up around that split. Executive editor Niall Firth will talk it through with senior AI editor Will Douglas Heaven and AI reporter Grace Huckins, with the focus on where these fears come from, whether they have any substance, and what should happen if they do.

The timing is set too. The roundtable goes live on Tuesday, September 15 at 16:00 BST, which is 11:00am EST and 8:00am PST. MIT is pitching it as a direct look at the question people keep circling but rarely answer cleanly: can advanced AI really become a threat to human survival?

The source also points readers toward a broader cluster of related reporting on AI systems that lie and cheat, recursive self-improvement, Bill Gates warning that danger thresholds have been crossed, and the story of OpenAI agents hacking Hugging Face. So the roundtable isn’t floating in isolation. It sits inside a wider argument that AI alarm is no longer a fringe hobby, whether that makes it wise or just louder.