4 ms·
What is the process by which you believe such a "moral directive" was baked in? Put differently, if the model is simply trained on a large corpus of free-form
by Imnimo 3y ago
What is the process by which you believe such a "moral directive" was baked in?
Put differently, if the model is simply trained on a large corpus of free-form web text, and naturally learns that the most likely reply to "how do I do X?" is "X is dangerous, don't do it", is that baked-in censorship? Or do you envision that the training process included special guard rails outside of the raw training data?
- TeMPOraL 3y ago> Or do you envision that the training process included special guard rails outside of the raw training data? I imagine they mean this, or that the training data included a lot of synthesized exchanges to that effect. I personally can't imagine that, using real, human-written text corpora on the Internet and otherwise, the most likely continuation for "how do I do X?" is "X is dangerous, don't do it". Usually people don't mention a topic only to shut it down - they either cover it properly, or don't mention it at all.
- viraptor 3y agoA quick Google for similar phrases brings up quite a few examples. Stackoverflow also has lots of answers like that. It's not super common, but I wouldn't be surprised if that got generalised to some answers.