Most AI vendors don't benchmark for reliability. A new benchmark from Princeton researchers does.

The technology used to predict sand and dust storm (SDS) severity has for decades systematically overestimated when and where ...

A new technique estimates the reliability of a self-supervised foundation model, like those that power ChatGPT, without the need to know what task that model will be deployed on later. Foundation ...

Last year, I participated in a roundtable discussion on artificial intelligence at Fluke Reliability’s Thought Leadership Day ...