5 ms·
AI Helped Me Tackle One of Math's Greatest Mysteries
- kilokilo22 2y ago[dead]
- kilokilo22 2y ago[flagged]
- DoctorOetker 2y agoSection 5.6.1: Calculation of the hypothenuses: a superfluous 4p appears on the right hand side for T1 and T2, LHS != RHS for T as well LHS != RHS Section 5.6.2: first line of equations: LHS != RHS second line of equations: LHS != middle expression != RHS ... giving up
- Circled 2y agothanks for the input. I will look into it. however, chapter 6ff tackles it with another approach. what you think about that?
- DoctorOetker 2y agoAre you the author of this paper? the mismatching LHS and RHS would be detected by 12y olds. Perhaps you are this young, in which case I approve of your attempts. Otherwise you are piping AI nonsense to the crowd which I disapprove of. So I assume you mean "Alternative formulation of the proof" in section 6: This relies on Lemma 15, which is right after where I gave up after about 3 basic errors. So let me continue. Your attempt to prove Lemma 15 relies on multiple steps, the first of which states: Step 1: the sequence a_n generates numbers of the Mersenne Prime form, where a_1 = 3 (and hence p=2), and where a_n+1 = a_n squared - 2. While 3 is indeed a Mersenne prime, the next number will be 9 - 2 = 7, still fine, but the third number in your sequence is 7 squared - 2 = 49 - 2 = 47, which is not equal to any power of 2 decremented. since Lemma 15 states false things, and your alternative proof relies on it, there is no alternative proof.
- Circled 2y agothanks for the input. indeed very kind of you to take the time. I guess it cannot be fixed then? actually the whole idea is to solve it with AI solely as an experiment. and of course human feedback is allowed. so since that was outlined clearly i am not piping nonsense to the crowd. like a turing test maybe...congrats, you passed ;) but i did take it offline since there is no value...
- DoctorOetker 2y agoOh, I certainly believe deep learning can solve mathematical problems at rates much higher than humans. But that would require using formal verifiers like metamath in the training loop. https://us.metamath.org/ https://us.metamath.org/ BTW if you found the experience slightly embarassing, get acquainted with metamath, and its interactive prover "mmj2", and the "set.mm" database. everything you prove in mmj2 is guaranteed to be correct (as long as you dont add axioms)
- nujsii 2y ago[flagged]
- Circled 2y agonot sure what is the link here. you seem to enjoy showing your intelligence. that is ok. if you create value in the world like that, you are doing the right thing