Andrea Paone

CEO at Paone's Army! Un pazzo con idee completamente fuori luogo. Scherzo, soltanto un normale cittadino.

[Andrea Paone @Andrew3000s]

Been trying Muse AI and it genuinely feels more like a personal assistant than a chatbot. It understands what you want and actually gets things done for you.

If you want to try it, use my invite code to get 1 billion Muse tokens that never expire:

ZK3S2E
https://t.co/0QbiYpG78t

[Andrea Paone @Andrew3000s]

@thomasfbloom @Liam06972452 @AcerFur Agree

[Andrea Paone @Andrew3000s]

@LizardWizardBTC The only reason I was sensitive is that another account reproduced my exact graph, graph6, certificate and matching checksum, then presented it in a different template as new

[Andrea Paone @Andrew3000s]

@LizardWizardBTC Yes, same local calendar date. And honestly, keep exploring. Turning these results into properly checked papers matters, because published work becomes part of the public record and knowledge and can be built on by anyone

[Andrea Paone @Andrew3000s]

@LizardWizardBTC Just to be clear, I’m not upset with you or accusing you of anything. I meant it as a friendly methodological exchange. I’m dealing separately with another case involving an exact reproduction of my material, so I was sensitive to the issue. Your case is different

[Andrea Paone @Andrew3000s]

@LizardWizardBTC The public Zenodo record went online at 22:29 UTC on July 22, with the DOI registered seconds later. Your tweet was at 03:08 UTC on July 23, about 4h39 later.

[Andrea Paone @Andrew3000s]

@LizardWizardBTC No worries, I wasn’t accusing you of copying. It was just a methodological tip about checking prior art. All clear

[Andrea Paone @Andrew3000s]

@LizardWizardBTC One more suggestion for the future. using AI in mathematical research is not a problem. The author must understand the argument, take responsibility for every claim, verify the mathematics independently, and check prior art. i hope it helped

[Andrea Paone @Andrew3000s]

@LizardWizardBTC I’m simply giving you advice before making novelty claims in a scientific context. Check the prior art first. It will help if you plan to publish.

[Andrea Paone @Andrew3000s]

@LizardWizardBTC I believe you may not have known. But that is exactly the issue: when GPT is asked to solve a “random open conjecture,” it may surface an already public result. The user still has to check prior art before presenting the output as a discovery.

[Andrea Paone @Andrew3000s]

@LizardWizardBTC This is exactly the problem with using AI without understanding the mathematics. Chen/Li disproved a different, general inertia inequality. They did not settle Conjecture 4.12 on line graphs. A paper is more than matching “graph inertia” in a title.

[Andrea Paone @Andrew3000s]

@LizardWizardBTC Chen/Li is a different result: it disproves the general inertia inequality, not Conjecture 4.12 on line graphs.

[Andrea Paone @Andrew3000s]

@hankakuspace7 @DmitryRybin1 @hankakuspace7 Your report reproduces my exact C5-C4-C5 graph, graph6 string, inertia and characteristic-polynomial certificate, published 4h46 earlier. AI use does not remove the duty to verify sources and cite prior work. Please correct the report.

[Andrea Paone @Andrew3000s]

@hankakuspace7 @DmitryRybin1 On 22 July 2026 the first connected counterexample to the Akbari linegraph inertia conjecture was already public: https://t.co/knbm2DhaGY
Presenting the same type of result the next day without citing the existing preprint is not how academic priority works.

[Andrea Paone @Andrew3000s]

@LizardWizardBTC @LizardWizardBTC To clarify: your graph is different from mine. However, the conjecture had already been publicly disproved on 22 July 2026. Any later presentation should acknowledge the earlier public result: https://t.co/AZ2PyJj2yp

[Andrea Paone @Andrew3000s]

@LizardWizardBTC On 22 July 2026 the first connected counterexample to the Akbari linegraph inertia conjecture was already public: https://t.co/knbm2DhaGY
Presenting the same type of result the next day without citing the existing preprint is not how academic priority works.

[Andrea Paone @Andrew3000s]

@__alpoge__ Hi Levent, was the Jacobian counterexample found in a single Fable conversation, or through an agentic Claude Code setup with repeated code/CAS experiments? Will you publish the prompt, transcript or search logs? I’d love to understand the actual discovery process.

[Andrea Paone @Andrew3000s]

@ProtonMail @Monero_XMR_1 Yes but at least give an option to enable it!