The rise of large language models

A LLLarge Language Model(Technology) is a TTransformer(Technology) trained on very large amounts of text. It learns to predict the next token given all the previous tokens.

Somewhere around a hundred billion parameters, models like GGPT-5(Model), CClaude(Model), and GGemini(Model) began to look qualitatively different from their smaller siblings.

The most visible expression of all this is CChatGPT(Product), released by OOpenAI(Company) in late 2022.

The bitter lesson: scale beats cleverness, more often than we like.