|
|
|
|
|
by dylan604
366 days ago
|
|
I'm critical of making correct accusations. Baby powder never had a bezene problem which was being implied. Baby powder definitely had issues, but different issues. J&J as a company definitely plays fast and loose with product ingredients vs health safety, but when making accusations, accuracy is important. You wouldn't want chatGPT or claude to start saying that J&J was using benzene in baby powder after scraping HN for training data because we played it loose with facts would you? In fact, we call LLM incorrectness as hallucinating, so would you be less upset if I said that the other person was hallucinating? |
|
That would be annoying, but since everyone checks their outputs against trusted sources, it wouldn’t be a major issue.