Estimating

There's a finding in AI research called emergent misalignment. Researchers took an aligned language model and fine-tuned it on one narrow thing: examples of insecure computer code. Bad code, nothing else. No malice anywhere in the training data. The model didn't just start writing bad code. It started giving harmful advice on subjects that had … Continue reading Estimating

What I Didn’t Say

In January of this year I used an AI tool to help me edit my novel. The chapter that came back wasn't mine. Same characters, different scene. Events I hadn't written, in a book I'd already spent a hundred and five thousand words on. I got about halfway down the page before I stopped and … Continue reading What I Didn’t Say