GPT-4 doesn't need training to learn a new language. You can just define it in the prompt. But if training is used, I think you would want to define a full sel

GPT-4 doesn't need training to learn a new language. You can just define it in the prompt. But if training is used, I think you would want to define a full self-contained platform. That way you know the LLM knows everything it needs to translate the user's requirements into code. A group like DeepMind might go as far as to integrate code execution into their system somehow. Maybe it's like an LLM but trained to iteratively write code for some self-contained platform. Or there could be a full platform designed so the specification fits well-within a context window. Something like K or oK is very compact. You could give it a library to save data and display some UI stuff and define the whole thing in a context window maybe.