Now just need a plausibility checking AI to vet GPT-3 outputs so the author does... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		imlina on July 19, 2020 \| parent \| context \| favorite \| on: OpenAI's GPT-3 may be the biggest thing since Bitc... Now just need a plausibility checking AI to vet GPT-3 outputs so the author doesn’t have to read through 10 versions of his potential blog post.

xsmasher on July 20, 2020 [–]

That's how generative adversarial networks work, right? One neural net to create, another to check if it matches the training data.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact