Artificial Intelligence
Innocent-looking AI reasoning can make bad behavior harder to catch
AI safety monitoring can fail when an AI’s reasoning is the main clue that something has gone wrong, new research suggests.
Artificial Intelligence
More Stories in Artificial Intelligence
- Artificial Intelligence
Is AI making us dumber? Maybe not. But our skills are at risk
Studies suggest AI can weaken learning when it replaces effort, but tools that guide rather than answer may help people keep their skills.
- Artificial Intelligence
Struggling with a hard life choice? AI future selves have tips
A chat with aged avatars of future selves can guide people paralyzed by tough decisions. The question is what information to feed the bots.
- Artificial Intelligence
AI ‘ghosts’ can comfort mourners — even when the bots get the facts wrong
A small study suggests emotional fit, not perfect accuracy, may make chatbot versions of dead loved ones feel real.
- Artificial Intelligence
AI is not ready to fly solo in space
In sci-fi, AI can navigate the unknowns and — ideally — keep human travelers safe. But it’s not intelligent enough to do that yet.
- Artificial Intelligence
AI tools meant to vet science are surprisingly easy to fool
The gold standard of scientific review, peer review by researchers’ colleagues, is in crisis. AI might offer a solution but has problems of its own.
- Artificial Intelligence
How big a cybersecurity threat are the latest AI models, really?
New AI models are accelerating the game of cat-and-mouse as cybersecurity experts try to keep ahead of would-be hackers. An AI expert explains the risks.
- Health & Medicine
A blood test for dementia may tell you if you have more than one type
AI helped researchers develop an experimental blood test that might let doctors diagnose overlapping dementias.
- Math
AI cracked an Erdős math problem. Now experts want guardrails
The result is correct but challenges core norms of mathematics: checking proofs, crediting ideas and keeping research open to everyone.
- Artificial Intelligence
AI bots ignore evidence. Can we trust them with science?
Scientists rethink their ideas after experiments. AI agents struggle to learn from evidence and recognize when an idea is obviously incorrect.
