Decrypt

There's a Benchmark Test That Measures AI 'Bullshit'—Most Models Fail

BullshitBench tests whether AI models can detect nonsensical questions—or if they'll confidently answer them anyway. The ...

I Wrote A Resume For A $180,000 Job Using ChatGPT 5.4 To Test The Hype

How do you write a resume using AI? I tested ChatGPT 5.4 and compared it to ChatGPT 5.2, to create an executive resume for a project director job. Here's what happened.
PublicTechnology

PAC digs into government’s ‘curiously specific’ claims of AI benefits

In a distinctly sceptical missive, the head of the parliamentary committee has warned top officials that recent research ...

I tried GPT-5.4, and most answers were really good - but a few had me concerned

I tried GPT-5.4, and most answers were really good - but a few had me concerned ...