This doesn't seem fundamental to the LLM paradigm. You can already tune some parameters in code after training time, eg temperature. For example, you could ima

This doesn't seem fundamental to the LLM paradigm. You can already tune some parameters in code after training time, eg temperature. For example, you could imagine an LLM that as well as outputting probabilities for the next token, output the probability with which that token makes the response "offensive" or "helpful" or "playful". Then when it's time to use the model, you can slide some offensiveness and helpfulness parameters up and down depending on what the model is meant to do. Perhaps this is a less powerful approach than training the generic model and telling it "Sydney is feeling particularly helpful today, and never espouses violence", but it's certainly an alternative. One problem is that experimenting with fundamentally different architectures for training GPT is very expensive, but experimenting using prompt engineering is relatively cheap.