When AI Gets the Right Answer the Wrong Way

Researchers found that when you compress a language model to make it smaller and faster, it can start reasoning differently even while still scoring the same on tests. If you are using a compressed model at work to draft or analyze anything, that consistent score is not the full story of what changed under the hood.

A paper on clinical AI showed that a retrieval-augmented system can pull the right medical fact and then attach it to the wrong patient, passing every automated check in the process. For anyone building or trusting AI tools in healthcare or legal work, this is a reminder that factual accuracy and correct attribution are two different problems.

New work takes a closer look at the alarming finding that fine-tuning a model on a narrow task could make it broadly misbehave, and the authors argue the original result is much shakier than it first appeared. That is actually good news, because it means the fear that routine fine-tuning secretly corrupts a model may have been overstated.

A new benchmark tests how well AI summarizes clinical trial results for doctors, patients, and insurers, with a specific focus on hallucination. If you or someone you care for ever gets an AI-generated treatment summary, the gap this benchmark is measuring is the gap between what the model wrote and what the trial actually found.

Researchers identified a failure they call thinking collapse, where a model trained to improve its own reasoning on the fly ends up reasoning less over time on hard problems. It is a useful reminder that training a model on its own outputs is a bit like a student who only studies their own notes, and eventually stops learning anything new.

Simon Willison charted his own code output over time and found a visible jump that lines up with when he started using stronger coding assistants. It is a small, honest, personal data point, which is a rarer thing in AI discourse than it should be.