ringleader

[ringleader @mljumaru]

@wm__meng @BunzaiM @provisionalidea The grade schooler trusts a process: independent experts who check the proof and answer for it. A smarter prover doesn't replace that; capability isn't verification, AGI or not. Muscle's output is the product. In maths the product is understanding, which can't be outsourced.

[ringleader @mljumaru]

@wm__meng Fable says: “He objects to a monopoly on knowledge, then proposes a literature written by agents, for agents, and reviewed by agents. That hands the monopoly to whoever owns the agents. And trustees who can't read the books aren't governing anything.”

[ringleader @mljumaru]

@BunzaiM @provisionalidea Also, since you’re fascinated by AI’s myriad of capabilities, here’s some of Fable’s takeaways from our exchange. You’ll probably disagree with it yet will trust any proof it gives you of Riemann’s Hypothesis at face value? Food for thought.

[ringleader @mljumaru]

@BunzaiM @provisionalidea You've just agreed that checking a proof means learning the mathematics. That's all I said. And nobody is kept out: the papers are free on arXiv and anyone can learn. Asking for results to be verified before they're accepted isn't arrogance or gatekeeping. It's science.

[ringleader @mljumaru]

@BunzaiM @provisionalidea You’re literally describing becoming a mathematician at that point, the idea is self defeating. For you to understand every step in a proof like Wiles’ or Chen’s requires knowing maths at that level, not just having it half-explained and half-blindly trusting an LLM.

[ringleader @mljumaru]

@haider1 It is the other way around, it is childish to believe your AI generated mathematics are mathematics at all, you’re not able to validate or falsify them yourself, all you have to show for them are maybe hallucinated lines of Lean. Mathematics can’t be done w/o mathematicians.

[ringleader @mljumaru]

@BunzaiM @provisionalidea It’s not that simple, the mathematics involved in a potential proof of the Riemann Hypothesis are no long division. Even existing proofs like Wiles Theorem cannot simply be explained to people outside mathematics. Your outlook is just naive.

[ringleader @mljumaru]

@BunzaiM @provisionalidea Except if GPT or Claude prove the Riemann Hypothesis, you have no way of reliably confirming such a proof is correct. The whole point of science being falsifiable breaks down if no one capable of doing so is involved, no matter how many lines of Lean an LLM can hallucinate.

[ringleader @mljumaru]

@EERandomness @provisionalidea Wrong, we’ve had the arxiv since 1991, it literally makes no sense to use git for anything other than code and its documentation

[ringleader @mljumaru]

@doomslide @__alpoge__ I doubt an internet nobody’s claim that Levent is a ‘collosal disgrace’ is much relevant. And it is clear that ‘normal pure math collaboration with ai’ makes reference to working on results *literally using ai*, but personally and not as company projects by OAI or Anthropic.

[ringleader @mljumaru]

@Dancinchica @scopely @ScopelyExplore The whole point of the City Safari is to explore the city, go to the booths for the stamp rally, do the in-person activities related to the event and meet people. A remote make up makes no sense.

[ringleader @mljumaru]

@resellingnews @digimintcoco

[ringleader @mljumaru]

@Asdf79873607 In Peru it’s the same, the spawns were nowhere yesterday

[ringleader @mljumaru]

@cuidado_obra Si Curwen habla puras cojudeces y se retiró de la Ruiz después de un par de semestres por burro

[ringleader @mljumaru]

@ScopelyExplore Where is Wiglett? Spawn rates are as trash as ever, where’s the boost?

[ringleader @mljumaru]

@PokemonGoApp Where is Wiglett? Spawn rates are as trash as ever, where’s the boost?

[ringleader @mljumaru]

@Dexerto Light work

[ringleader @mljumaru]

@sahelmulachela @tw3r1ckx @LechedeDios1 @PGSatelite How many Mew do you have? When are you expecting your third? Only two (one normal, one shiny) since 2016? When is the third coming, 2031?

[ringleader @mljumaru]

@GavinXCIX @LechedeDios1 @PGSatelite GM Machamp? It’s rarer than Mewtwo and the best fighting type max attacker… the odds of already having a hundo Mewtwo by now must be like 100x that of having a hundo GMAX Machamp. I’ve caught hundreds of Mewtwo (w/ higher iv floor during the weekend) and less than 30 GM Machamp.

[ringleader @mljumaru]

@sahelmulachela @tw3r1ckx @LechedeDios1 @PGSatelite Notice how you got only 4 of those, you could have raided thousands of mewtwo in the same timespan you caught those 4 Shaymin. Also, if they put the masterwork research in store ever again, it’ll likely give you candy rather than another shiny encounter, same as with the rest.