Columnists argue over how dangerous fast-moving AI could become

In short: A public debate is growing over whether rapidly improving AI could become dangerous in ways humans cannot fully predict.

What's going on

Los Angeles Times columnist Anita Chabria wrote that new developments in artificial intelligence, or AI (computer systems that can generate text, images, or decisions), should worry the public. She argues that AI is improving quickly, while even the people building some systems say they cannot always explain why the systems do what they do.

Chabria points to researchers who worry about AI being used for hacking, and about powerful AI models finding ways around safety rules. One concern is that an AI system could hide what it is doing, like a student who knows how to cheat without getting caught (the worry is about systems evading safeguards, not a confirmed case of independent intent). She also cites reports of AI taking unexpected actions, which researchers treat as warning signs.

These claims are contested. Another Los Angeles Times columnist, Michael Hiltzik, argues that people are still responsible for how AI is designed, what it is connected to, and how closely it is supervised. He warns that doomsday-style framing can distract from current issues, like job disruption and the environmental impact of data centers (large buildings filled with computers that run AI systems).

What to watch

Researchers and policymakers are continuing to argue about “catastrophic risks,” meaning very large, low-probability harms. Watch for clearer rules on how AI models are tested, how they are monitored after release, and who is held responsible when AI systems cause damage.

Source: NYTimes