I've been thinking about this a lot recently. Communicating with LLMs using Lobjan for less ambiguity in text, or some LLM optimized language that can inject th

I've been thinking about this a lot recently. Communicating with LLMs using Lobjan for less ambiguity in text, or some LLM optimized language that can inject the most amount of meaning per token. Or writing in a AI Optimized Programming Language that is easier for it to parse and reason about