I was pulled over for having a non-obstructive frame on my license plate. The officer said they interfere with the red light cameras. He then presented me with a screwdriver and gave me the option of getting a ticket or taking it off. I took the screwdriver and he watched me take it off. I lost a freedom due to a shitty ml model.
Tried all ten with claude, then had codex take a loook at the work -- codex thinks number 7 has the lowest chance of being correct, a 1 out of 10 rating. None of them were higher than 7/10 chance of being right so far as done by claude opus 4.6 and evaluated by codex 5.3 highest.
I don't think either of these are the best choices for this. Chatgpt 5.2 pro and gemini 3 pro deep thinking I believe are the strongest LLMs at "pure thought", i.e. things like mathematical reasoning.
Then nice try I guess, but there is nothing remotely new about any of this. You're not going to find a new perspective on something in math using a completely routine application of a well-known theorem. You'll need to either work really hard or come up with a genuinely new idea, and probably both.
I just found it interesting that certain problems in number theory could be rephrased as problems about cyclic groups. Maybe it could potentially make some easier to solve.
reply