The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Claude AI is a little bitch
| chilmata | 09/05/25 |
| .,.,.;.,..,..,.,:.,:,..,..,::,..,:,.,.:,..:.,:.:, | 09/05/25 |
| chilmata | 09/05/25 |
Poast new message in this thread
Date: September 5th, 2025 1:40 PM Author: chilmata A couple weeks ago it refused to draft me a simple legal brief. It literally told me I should do it myself since I'm a lawyer. So I had ChatGPT do it without any pushback.
Last night I'm in the process of refining my groundbreaking theory, The Law of Competing Probabilities, to apply it to exploiting inefficiencies in the sports betting market and Claude goes, "Do you really think sports betting is the best use of your time. Maybe you should put the AIs down for a bit and focus on something more practical."
Fuck Claude. I had Gemini do a deep research dive and he validated the scientific and mathmatical basis for my theory as the other AIs have done.
Claude isn't saying my theory isn't plausible, its telling me to get some fresh air. WTF is that shit? That's not why I'm paying it $30 a month. Canceled.
The difference between me and the guy in the NY Times article who ChatGPT convinced he was a genius and invented a new math theory: Gemini falsified his theory and validated mine. |
Date: September 5th, 2025 1:42 PM Author: .,.,.;.,..,..,.,:.,:,..,..,::,..,:,.,.:,..:.,:.:, State your theory here so we can validate it (http://www.autoadmit.com/thread.php?thread_id=5770126&forum_id=2"#49238910) |
 |
Date: September 5th, 2025 2:08 PM Author: chilmata The Law of Conflicting Probabilities basically states that Regression to the Mean in sports streaks can be envisioned as competing probability vectors that produce predictable outcomes that are mispriced by the sports betting market line and that their likely outcome is amplified by the the size of the total perceived audience (Spotlight Effect).
The longer a streak goes on, the more likely it is that it will "snap back" due to pressure from "statistical gravity."
This is distinguishable from gambler's fallacy.
The most basic example of the competing probability vectors is the "batting average" vector competing with the "pitcher's ERA" vector. Those vectors will compete to produce an outcome (e.g. a home run or a strikeout).
That's an oversimplification.
|