3 ms·
Language model researchers, like social media developers before them, have built this tech under the presumption that it will be applied only with positive inte
by kibibu 4y ago
Language model researchers, like social media developers before them, have built this tech under the presumption that it will be applied only with positive intent.
GPT-4chan is about the least worst thing somebody could have done with it.
- heresie-dabord 4y agoYes, much worse is possible. Any good AI/ML language-model implementation will produce results indistinguishable from those of a human... And that is (even in the best of current civilisation) a wide range of capacity and behaviours, the folly of which is usually limited only by societal pressure.