4 ms·
every 3rd sentsnce is "the model was not trained on data from meta's products"
by md_rumpf 2y ago
every 3rd sentsnce is "the model was not trained on data from meta's products"
- keyle 2y agoThat makes sense. You probably don't want to train your LLM against your uncle's dubious claims about the government, flat earthers and the likes content :)
- esafak 2y agoAt least they're clear about provenance, unlike OpenAI. https://www.reddit.com/r/ChatGPT/comments/1bfa7s3/openai_cto_mira_murati_confirms_that_the_video/ https://www.reddit.com/r/ChatGPT/comments/1bfa7s3/openai_cto...