All the Sovcit logic stuff is bullshit anyway, it just sounds like something that might be real. So build an LLM and feed it all the Sovcit shit you can find. Then ask it a question that contradicts a Sovcit position. Yes its going to be a garbage answer it returns, but it will sound like something a Sovcit would say.
Sovcit: “You can’t take my car because I issued an international statement of payment with receipt on my account!”
LLM answer: “Oh, I’m sorry. International statements of payment are only valid in registered districts of independence. You clearly haven’t executed a notice Promissory Estoppel, and its too late now so I have to take the car.”
Could this be a use great use case of an LLM?
All the Sovcit logic stuff is bullshit anyway, it just sounds like something that might be real. So build an LLM and feed it all the Sovcit shit you can find. Then ask it a question that contradicts a Sovcit position. Yes its going to be a garbage answer it returns, but it will sound like something a Sovcit would say.
Sovcit: “You can’t take my car because I issued an international statement of payment with receipt on my account!”
LLM answer: “Oh, I’m sorry. International statements of payment are only valid in registered districts of independence. You clearly haven’t executed a notice Promissory Estoppel, and its too late now so I have to take the car.”
Like lemmy bot… but more stupid. I like it.
It’d be funny to build a bot to get in their replies, but doing so would help feed the problem. But I do like your approach.
Damn that’s such a good idea lul.