Assuming this pans out, every other professional linguist in the world has had the option to use Claude or other LLMs, but has not solved this problem, despite the incentives for doing so. It stands to reason the human is adding crucial value.

I drilled down on this with Tom. He thinks that it might not have happened without Claude Code, but Claude was used to organize all of the symbols, and to run I think it was 100,000 simulations to assess whether or not he had an actual insight, or if he just randomly got lucky. Claude did NOT crack the code. Significant supporting role though.

So Claude Code was used to generate software that ran simulations? I don't think LLMs in and of themselves can execute simulations, esp. a specific, non-single digit count like 100k.

I don't know if the agent ran the simulations or if the agent built software that ran the simulations. But Claude was used to run the simulations.

I feel like that is extremely relevant information. An LLM running "simulations" is nonsense, using an agent (apparently Claude Code) to write simulation software is more realistic. I'm sure you could clarify this since you have a relationship with the person.

Curious to see where this goes. Hopefully an update is posted here when it's all said and done :)