Honestly I’m pretty tired of supporting nostalgebraist-autoresponder

GPT-4 prediction: it won’t be useful

The scikit-learn cargo cults

Meta-Post on Meta-Learning

Talk about “AI” in the press these days tends to conflate two things

Gary Marcus's Kafkaesque Critique of GPT-3

On “Learning to Summarize”

GPT-3 and Scaling Trends

Hosting a GPT-2 autoresponder bot

Tensorflow User Experience

GPT-2 is not as dangerous as OpenAI thought it might be

Trees are harlequins, words are harlequins – the transformer “explained”?