I say this as someone who's found LLMs incredibly beneficial.

Is this a way to increase token burn?

I thought we covered this with Claude's C compiler. What changed?

My initial reaction was that this is tokenmaxxing disguised as a product.