MIT Technology Review to host roundtable on AI extinction fears

The event, Roundtables: Will AI really kill us all?, will feature Niall Firth, MIT Technology Review’s executive editor, Will Douglas Heaven, senior AI editor, and Grace Huckins, AI reporter. According to MIT Technology Review, the conversation will unpack AI extinction fears: where they come from, whether they hold any water and, if so, what should be done.

MIT Technology Review said employees at the world’s leading AI labs have said there is a real possibility that advanced AI could destroy humanity. The roundtable will ask whether those warnings are right or whether they amount to more scaremongering and hype.

The publication listed related stories on AI agents that lie and cheat to reach their goals; the possibility that AI’s recursive self-improvement may not come as quickly as some expect; Bill Gates saying we’ve passed AI’s danger thresholds; and the inside story on why OpenAI agents hacked Hugging Face.

MIT Technology Review’s broader AI coverage includes a report on a fundamental flaw that leaves large language models strikingly vulnerable to attack, making it easy to trick them into doing things they should not do, such as telling someone how to sabotage an aircraft’s navigation system. Another piece reports that AI is more likely than humans to form biases when hiring, not only learning stereotypes from training data but also cooking up new ones.