4 ms·
These models are trained on lots and lots of science fiction stories, of course they know how to autocomplete questions about AI ethics in ominous ways
by ajayyy 4y ago
These models are trained on lots and lots of science fiction stories, of course they know how to autocomplete questions about AI ethics in ominous ways
- Gigachad 4y agoI think the more interesting point is that even if we dismiss this as not real AI because it isn’t that advanced yet, at what point would we have achieved real AI? How would we identify it and categorise basic language models from actual thinking?