Ask HN: My friend says he has an AI optimized language

1 points by samweb3 257 days ago | 2 comments

A friend built https://memelang.net/08/ which claims "Memelang is an AI-optimized query language that significantly reduces token count and model size for LLM RAG. Copy-and-paste the Python spec below into your LLM."

Is it possible that this language would make AI agents better at retrieving data either in terms of cost or accuracy? For example, might an AI Agent calling tools be more efficient than just using standard SQL or similar? How would you design a test or set of tests to figure out the benefits?

I would like to help him prove the cost or accuracy benefits. Thank you for any advice!

anonymoushn 257 days ago |

These example queries really do take fewer tokens than the SQL queries, and the syntax is apparently designed to minimize token count.

My impression is that letting the model write SQL queries is not a typical baseline for RAG. It's unlikely that a company could substantially benefit from this reduction, because the cost of their internal LLM usage is probably mostly tokens that are not RAG queries or tabular responses. They also probably cannot fall back to a cheaper model because the "minimum acceptable level of intelligence" of the model is likely chosen for tougher constraints than the ability to generate correct RAG queries.

I don't think this is a reasonable thing to make a business around, because the claimed benefits mostly do not exist.

Also, 9 months ago you posted that you were building it, I suppose? https://news.ycombinator.com/item?id=42169086

samweb3 256 days ago | |

Yes, helping to build it really through experimentation and looking for use cases that make sense. Thanks for the comment!