The problem with language models is that while they are very good at giving you the information you want, they don’t tell you what to look for in the first place. You have to figure that out yourself, and you have to learn how to be good at asking for it. This is the metaskill that has become crucial over the last few decades, when our access to... See more

What’s ChatGPT really bad at? Being counterintuitive and surprising us. Large language models (LLMs) are trained on the mass output of humanity to optimize for what is expected. As a result, ChatGPT routinely let’s us down when it comes to getting something we didn’t expect. If you start paying attention to the music, books, movies, speeches,... See more