Yeah, but the hard part is filtering. It’s pretty easy to scrape a massive amount. Turning it into quality training data is the trick. bmk was the magician the

Yeah, but the hard part is filtering. It’s pretty easy to scrape a massive amount. Turning it into quality training data is the trick. bmk was the magician there. https://twitter.com/nabla_theta?s=21&t=Gt6YrATJHnmY046MdzhYDg https://twitter.com/nabla_theta?s=21&t=Gt6YrATJHnmY046MdzhYD...