The reason these GPT large language models stand out is, we "generate text" based on the "input text" we are trained on too. But these can train on orders of ma

The reason these GPT large language models stand out is, we "generate text" based on the "input text" we are trained on too. But these can train on orders of magnitude more text than we will ever ingest in our entire lifetimes.