I gotta say that the new @openai language model API is surprisingly good. If you can define a language task through a few (<10) examples, it seems to just work.
No rigorous tests yet, but through the GUI it's downright fun.
3
3
54


