Achim Mueller (@effoderia) header
Achim Mueller (@effoderia) profile picture

Computer Science (Diplom-Informatiker), Journalist (FJS). Private Account. Cryptography veteran, Patent holder, BTC Maxi. Threema-ID https://xitter.gq/t.co/VkZ8o1kqIE

Berlin, Deutschland
Joined May 2026
Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)Photo by Achim Mueller (@effoderia)
Achim Mueller retweeted
🇩🇪👁️ Jeudi 12 novembre le Bundestag a adopté la grande réforme de la douane contre la criminalité financière Dedans il y a un paragraphe (le § 18a) dont personne ne parle... La douane pourra collecter sur internet des données publiques contenant tes traits biométriques, ton visage sur une photo, et les passer dans une comparaison automatisée avec ses fichiers Si la douane n'y arrive pas elle-même la comparaison peut être confiée à un organisme privé. Et même hors de l'UE au "nom de la sécurité nationale" en principe sur décision d'un juge Côté argent... Ce serait pour un bien de plus de 100 000€ à l'origine "douteuse" cryptos écrites noir sur blanc "peut être saisi 30 jours sans procédure pénale. Prolongation de 150 jours par un juge seul, sans audience, décision sans recours"... Et c'est à toi de rendre l'origine plausible ! Le Bundesrat doit encore voter pour une entrée prévue au 1er janvier 2027
7
49
3
132
14,432
Achim Mueller retweeted
🚨EXEMPLO: Após escândalo bancário no Paraguai, o presidente renunciou e, em menos de 72h, todos os diretores e o dono do Ueno Bank estão presos O dinheiro já está sendo devolvido para 98% dos correntistas.
273
3,354
159
19,748
322,961
Achim Mueller retweeted
Picture this. You have $50,000 saved at Paraguay’s biggest bank by number of customers. Monday night, the bank tells everyone it’s “operating normally.” Tuesday morning, the government takes it over. Your app is frozen. This week, you can get about $1,600. Later, up to about $37,000, on a date nobody has set. The other $13,000 gets in line with everyone else the bank owes. That’s Ueno Bank, right now, for nearly 3 million people. Eight months ago, JPMorgan and Bank of America were selling its bonds to investors around the world. Money in a bank isn’t sitting there waiting for you. It’s a loan to the bank. And when things go wrong, someone else decides when you get it back.
3
5
29
3,345
Achim Mueller retweeted
🚨 DEIN BITCOIN. DEINE WOHNADRESSE. DEINE STEUERNUMMER. ALLES IN EINER AKTE! Muss der Staat wirklich jeden Krypto-Kunden erfassen – auch ohne den geringsten Verdacht? Mit der EU-Richtlinie DAC8 wird genau das Realität! Seit 1. Jänner 2026 sammeln meldepflichtige Krypto-Börsen, Broker und Apps entsprechende Kundendaten und Transaktionen. Ab 2027 werden diese Informationen jährlich automatisch an die Finanzbehörden gemeldet und international ausgetauscht. 👉 Name und Wohnadresse 👉 Steueridentifikationsnummer 👉 Käufe, Verkäufe und weitere meldepflichtige Transaktionen Und das völlig unabhängig davon, ob jemand jemals etwas Illegales getan hat! ⚠️ WARUM IST DAS GEFÄHRLICH? Bitcoin ist ein digitaler Vermögenswert, der bei eigener Verwahrung unmittelbar übertragen werden kann. Geraten persönliche Daten samt Informationen über Kryptotransaktionen in falsche Hände, können sie Kriminellen helfen, potenziell vermögende Personen gezielt auszuwählen. Bitcoin Austria berichtet unter Berufung auf Frankreichs Innenministerium von 77 Entführungen und Erpressungen einschließlich Versuchen mit Kryptobezug allein im ersten Halbjahr 2026! Das ist kein theoretisches Sicherheitsrisiko. Natürlich müssen Steuerhinterziehung und Geldwäsche bekämpft werden. Aber rechtfertigt das die systematische Datensammlung über Millionen unbescholtene Bürger? Und wo ziehen wir die Grenze? Heute Kryptotransaktionen. Morgen weitere Vermögenswerte? ❓ Ist DAC8 ein weiterer Schritt in Richtung eines umfassenden europäischen Vermögensregisters? Ein solches Register wird durch DAC8 nicht geschaffen. Die Frage nach der zunehmenden Verknüpfung finanzieller Daten halte ich dennoch für berechtigt. 🇦🇹 @Bitcoin_at Verein WEHRT SICH! Bitcoin Austria hat eine Initiative gestartet, um die Rechtmäßigkeit dieser Datenerfassung überprüfen zu lassen. Eine erste Beschwerde wurde bereits bei der österreichischen Datenschutzbehörde eingebracht. Langfristiges Ziel ist eine gerichtliche Klärung bis hin zu den europäischen Höchstgerichten. 👏 Ich finde es wichtig, dass diese Fragen öffentlich und rechtlich geprüft werden! 🔗 Den Beitrag von Bitcoin Austria und die Initiative dac8.org mit allen Details findet ihr hier: bitcoin-austria.at/beitraege… Mich interessiert eure Meinung: Wo endet die legitime Steuerkontrolle und wo beginnt ein unverhältnismäßiger Eingriff in unsere finanzielle Privatsphäre? Beste Grüße Euer Matthias 💙 RETTE DEIN GELD – Finanzbildung zum Selbermachen. 👉 Folgt RETTE DEIN GELD für unabhängige Finanzbildung und lasst gerne ein Like da, wenn euch solche Informationen wichtig sind! #Bitcoin #DAC8 #Datenschutz
17
29
178
4,615
Achim Mueller retweeted
Peter Thiel's critique of European tech is sharper than most people quote it: "In the social democratic European societies, it's acceptable to be moderately successful. It's not acceptable to be wildly successful." That's not a cultural observation. It's a structural diagnosis. Look at the pattern: Skype sold to Microsoft. Spotify relocated to New York and listed in the US. DeepMind sold to Google before it could compound. Klarna — one of Europe's most valuable fintechs — spent years being investigated and regulated the moment it hit scale. The moment a European startup gets big enough to matter, the social contract shifts. Regulators, press, and public sentiment move from neutral to hostile. So founders learn to exit early — not because they lack ambition, but because the system taxes it. Europe produces world-class talent and early-stage innovation. It just doesn't retain the compounding. The constraint isn't capital. It's the ceiling culture places on acceptable ambition — and founders respond rationally to the incentives they're actually given.
147
534
101
3,816
472,770
Achim Mueller retweeted
„Die Linken sind die Faschisten des 21. Jahrhunderts“, sagt Wolfgang Büscher, Sprecher von der Arche Berlin, im EMMA-Interview. Warum er nach 25 Jahren Jugendarbeit Alarm schlägt. emma.de/artikel/faschisten-d…
114
989
70
3,701
58,399
‼️ EU governments approved a GDPR change to let companies train AI on Europeans' personal data without consent. It makes AI training a possible "legitimate interest", judged case by case. Parliament must still agree. Max Schrems calls it a "digital expropriation" of Europeans.
294
1,743
657
7,068
618,187
Achim Mueller retweeted
This is an incredible moment in this history of mathematics, computer science, physics & science with some of the hardest problems in human history being solved by AI over a period of weeks. Depending on your perspective, it's either a crisis, a triumph, or both. I don't think anyone knows for sure. We're all trying to make sense of it. I oscillate between fear/dread and optimism/excitement on a daily basis. The rate of progress is insane, including today's announcement from OpenAI of more problems solved. To help make sense of all of it, here's my conversation with three mathematicians about this moment. Each of them is an expert in the field of each of the biggest problems solved: Navier-Stokes, quasi-Riemann hypothesis, the unique games conjecture and complexity theory more generally. Ankur Moitra is a mathematician and theoretical computer scientist at MIT, visiting the Institute for Advanced Study in Princeton. Steven Brunton (@eigensteve) is a mathematician and fluid dynamics expert at UW. Alex Kontorovich (@AlexKontorovich) is a mathematician at Rutgers specializing in number theory.
327
808
110
6,296
509,464
Personne n’aurait imaginé qu’il se dresserait un jour contre le véritable mal ou qu’il s’aventurerait dans les recoins les plus sombres du monde humain. Pendant des années, Hollywood s’est moqué d’Ashton Kutcher. À leurs yeux, il n’était qu’un acteur de sitcoms, un beau gosse en couverture des magazines, le visage associé à des émissions de caméra cachée comme Punk’d. On le trouvait charmant et drôle, mais personne ne le prenait vraiment au sérieux. Et pourtant, ce malentendu s’est révélé être une bénédiction. En 2009, Ashton Kutcher a discrètement cofondé Thorn, une organisation dont l’objectif n’était ni d’organiser de somptueuses soirées sur les tapis rouges ni de donner des leçons de morale. Son ambition était radicale et directe : développer des technologies de pointe pour traquer les trafiquants d’enfants. Il n’y eut ni grande campagne médiatique ni opération de communication tapageuse. Il a simplement décidé d’agir là où tant d’autres détournaient le regard. Kutcher s’est plongé corps et âme dans les profondeurs obscures du monde numérique. Pendant des années, il a appris comment les prédateurs en ligne opèrent, comment ils communiquent, comment ils échangent des contenus abominables derrière des écrans fermés et comment ils échappent à la justice. Il ne s’est pas contenté de financer le projet. Il a travaillé directement avec des ingénieurs en informatique, des experts en cybersécurité et des services internationaux chargés de l’application de la loi. Ensemble, ils ont développé des outils d’intelligence artificielle capables d’analyser le Web, de détecter des signaux suspects, de cartographier des réseaux clandestins et d’identifier les victimes avant qu’il ne soit trop tard. Il n’était pas simplement une célébrité chargée de prononcer un discours soigneusement préparé. Il était au cœur même des opérations. Il s’asseyait avec les développeurs, élaborait des stratégies et analysait des données brutes. Il a assumé toute la lourde responsabilité de ce combat. En 2017, Ashton Kutcher s’est présenté devant le Sénat américain pour témoigner sur la traite des êtres humains. De nombreux élus s’attendaient à entendre un discours classique de célébrité, rempli de formules émouvantes et de mots soigneusement choisis pour frapper les esprits. Mais il leur a présenté des faits concrets, des données issues des systèmes informatiques et des cartes détaillées de réseaux criminels à l’échelle mondiale. Il s’est exprimé d’une voix calme et posée, dont la force dépassait celle des discours les plus tonitruants. « J’ai vu des choses qu’aucun être humain ne devrait jamais avoir à voir », a-t-il déclaré devant les sénateurs, la voix brisée par une émotion sincère. « J’ai regardé des images vidéo montrant un enfant du même âge que le mien se faire violenter sexuellement par un homme. Et lorsqu’on voit cela, on comprend que ne rien faire n’est plus une option. » Un silence profond s’est installé dans la salle. Ce n’était pas une mise en scène. C’était le poids accablant porté par un homme qui avait contemplé l’abîme et choisi de le combattre. Depuis, les logiciels développés par Thorn ont aidé les forces de l’ordre à identifier des centaines de milliers de victimes à travers le monde. Ils ont soutenu des opérations policières dans des dizaines de pays et permis de réduire certaines durées d’enquête, qui se comptaient autrefois en semaines, à quelques heures. Mais surtout, ces outils ont contribué à ramener de véritables enfants en lieu sûr, auprès de ceux qui pouvaient les protéger. Pourtant, on l’entend rarement s’en vanter dans les émissions de télévision en soirée. Ashton Kutcher ne publie pas de photos d’enfants secourus, ne cherche pas à tirer profit médiatiquement des arrestations et ne court pas après les applaudissements. Il choisit délibérément de rester dans l’ombre, laissant la technologie et les enquêteurs parler à sa place. Dans le monde du divertissement, il reste peut-être cet acteur décontracté que l’on associe aux sitcoms. Mais pour les prédateurs qui sévissent sur Internet, il représente un adversaire redoutable : un homme qui a retourné contre eux leurs propres outils numériques pour en faire des pièges. Il n’a pas incarné un super-héros sur un écran de cinéma. Il est devenu quelque chose de bien plus puissant dans la vie réelle : un homme prêt à s’enfoncer dans les ténèbres pour permettre à d’autres d’en sortir et de retrouver la lumière. Aujourd’hui, quelque part dans le monde, un enfant dort paisiblement dans son propre lit, serrant contre lui son animal en peluche préféré, sans se douter des algorithmes complexes qui ont contribué à le sauver. Il ne connaît pas le nom d’Ashton Kutcher, et peut-être ne le connaîtra-t-il jamais. Mais parce qu’un homme a refusé de détourner les yeux, cet enfant a désormais la possibilité de vivre une histoire dont la fin sera heureuse. #AshtonKutcher #Thorn #ProtégeonsLesEnfants #LespoirEnAction #HérosDuQuotidien
225
1,634
69
6,367
152,125
Achim Mueller retweeted
A lighter and more flexible German industrial humanoid robot Germany’s NEURA Robotics has unveiled its 4NE1 Gen 4, featuring a slimmer design, a black textile outer layer and a 20% weight reduction over its predecessor. Based on the previous generation’s 80 kg weight, Gen 4 would weigh around 64 kg, though its exact weight remains unconfirmed. Its brain is Neuron AI, NEURA’s own Physical AI foundation model. It combines vision, touch, force, sound and language to help robots understand their surroundings and execute tasks. Through Neuraverse, learned skills can be shared across connected robots. NEURA is also building a 2,300-square-meter RoboGym with TUM to bridge the gap between training and deployment. Robots learn industrial skills through real-world data, human demonstrations and simulation... while NEURA advances real-world projects with partners such as Schaeffler,which plans to deploy thousands of its humanoids by 2035, and HD Hyundai Samho, where robots are being tested for shipbuilding.
77
146
28
635
42,116
Achim Mueller retweeted
A machine learning professor at Carnegie Mellon spends his days writing mathematical proofs about why deep learning works at all. He has no Twitter following. He runs no startup. He gives no TED talks. He has never been on a podcast. He has won one of the most prestigious early-career awards in American academic science and trained a generation of researchers who now work at OpenAI, Google, MIT, and Stanford. Most people scrolling AI Twitter have never heard his name. His name is Andrej Risteski. Here is the story, because the theoretical foundations of modern AI are being written by people the industry rarely talks about. Andrej did his PhD at Princeton in the Computer Science department under Sanjeev Arora. To anyone outside the theoretical ML world this name means little. Inside it, Arora is one of the most influential theoretical computer scientists alive. He runs the theoretical foundations of deep learning group at Princeton. He is the closest thing the field has to a school of thought, and Andrej trained directly inside it. After Princeton, Andrej moved to MIT as a Norbert Wiener Research Fellow, a position jointly held in the Applied Mathematics department and the Institute for Data, Systems, and Society. The Wiener Fellowship is one of the rare postdoctoral positions in the United States explicitly created for researchers working at the intersection of mathematics and machine learning. The math department does not give these out lightly. He joined Carnegie Mellon as an Assistant Professor in the Machine Learning Department. He has since been promoted to Associate Professor. His research focuses on the questions nobody writes viral threads about. Why do generative models work. Why does self-supervised learning produce useful representations. When does out-of-distribution generalization actually happen. What are the mathematical guarantees underneath the systems the industry shipped to a billion people without proving anything formal about. The work is dense. The papers have titles most engineers would skip. Provable Benefits of Score Matching. Sampling from Energy-Based Models with Score Matching. Theory of Self-Supervised Learning Through the Lens of Probabilistic Generative Models. The kind of papers that get cited heavily inside academia and almost never get a Twitter moment. He has won the NSF CAREER Award, an extremely competitive grant given annually to a small group of early-career US scientists deemed most likely to shape their field. The award itself comes with around half a million dollars in research funding over five years. His specific grant is titled "Theoretical Foundations of Modern Machine Learning Paradigms: Generative and Out-of-Distribution." He has also won an Amazon Research Award for work on causal and out-of-distribution learning. He is supported by additional NSF awards for foundational work on self-supervised learning. His collaborators are a who's-who of theoretical machine learning. Ankur Moitra at MIT. Rong Ge at Duke. Holden Lee at Johns Hopkins. Frederic Koehler at UChicago. Yu Bai at OpenAI. Pranjal Awasthi at Google. Moses Charikar at Stanford. Yoon Kim at MIT. The same names appear on his Google Scholar co-author list year after year. These are the people writing the theorems the industry is starting to rediscover when its models break in production. Andrej runs the Modern Paradigms in Generalization program at the Simons Institute at Berkeley as a visiting scientist. He has been there twice, once in 2021 on Computational Complexity of Statistical Inference, and again in 2024. The Simons Institute is the premier venue in the world for theoretical computer science. Most people who get invited once never get invited back. He does not have a public profile in the way the industry usually defines one. His personal website at andrejristeski .org is a flat page with no styling. His Google Scholar page lists papers. His CMU page lists an office number and an email. That is all. A Princeton-trained mathematician at Carnegie Mellon is quietly writing the theorems that explain why the most expensive systems in technology actually work. He just is not online about it.
8
43
3
292
26,480
BREAKING: 🇫🇷 BILLIONAIRE JACK DORSEY'S BITCHAT JUST BECAME THE MOST DOWNLOADED APP IN FRANCE THE APP THAT LETS YOU SEND #BITCOIN WITHOUT INTERNET IS NOW #1 ON THE APP STORE WHILE PROTESTS AND RIOTS HIT THE COUNTRY OFFLINE FREEDOM TECH IS WINNING 🔥
431
1,836
196
13,484
1,204,021
ANDREAS ANTONOPOULOS EXPLAINS HOW HE SECURES HIS #BITCOIN: "I WOULD CERTAINLY NOT PUT MY ENTIRE NET WORTH ON MY COMPUTER" "I ASSUME EVERYTHING I’M DOING IS MONITORED, INCLUDING MY PASSPHRASES, INCLUDING MY KEYS" "I BOOT ONTO A COMPLETELY BRAND NEW MACHINE OFF A CD" "I PRINT PAPER WALLETS, AND THOSE KEYS HAVE NEVER BEEN ONLINE" NOT YOUR KEYS, NOT YOUR BITCOIN 🔥
80
226
17
1,386
133,973
Achim Mueller retweeted
CLAUDE SOLVES 25-YEAR-OLD MYSTERY SURROUNDING NSA CRYPTOGRAPHIC STANDARDS Anthropic researchers report that Claude solved a decades-old mystery surrounding the origins of the NIST elliptic curves used in widely adopted cryptographic standards. For more than 25 years, cryptographers questioned how the NSA generated the mysterious seed values used to construct these curves, raising concerns that they could contain hidden backdoors. In under 30 minutes, Claude reportedly discovered that the seeds were generated using a simple 1990s pseudorandom number generator initialized with Unix timestamps. The discovery allowed researchers to reproduce the seeds for five major NIST curves, including P-256, and significantly weakens suspicions that their parameters were deliberately manipulated. However, some questions remain unanswered, including why certain candidate curves were skipped during selection. IMPORTANT FOR BITCOIN: Bitcoin uses ECDSA with the secp256k1 curve, not the NIST P-256 curve investigated here. The discovery does not reveal a vulnerability in Bitcoin’s cryptography or change its security assumptions. Anthropic says the findings are not a cryptographic break and require no action from users.
19
50
9
380
58,728
Achim Mueller retweeted
What does a superhuman level of mathematics look like? Here is one elementary example I could give to nonexperts. Daniel Kane and Terence Tao showed (arxiv.org/abs/1702.00912) that, given a collection X of sets, the total number of ways to write A as a disjoint union of B and C, with A, B, C from X (disjoint means non-overlapping), is at most cardinality of X to the power 1.72598…, and this power is sharp. To show this, they came up with an innocent-looking function B(x,y,z) = (xyz)^c of three variables, satisfying a certain arithmetic inequality and having prescribed boundary conditions. Proving that arithmetic inequality was really a headache. There was a question whether one could improve this bound, i.e., maybe add some logarithmic correction terms, and one could kind of imagine what should be done (see comments on terrytao.wordpress.com/2017/…), but it was not clear whether it would be easy to do. Now superintelligence improves this bound by introducing a certain monster function of three variables (see screenshot). Even if I had seen this function in my dreams, I would never have dared to verify that main arithmetic inequality. The candidate function produced by SI gives a bound matching the known lower-bound examples up to a constant factor, including the sharp logarithmic correction. All the details are here paatai.github.io/ai-proofs/b…
79
101
17
1,175
214,099
Achim Mueller retweeted
The story of the last Japanese soldier is well known throughout the world. The last German to surrender was Georg Gärtner. Georg Gärtner was captured in 1943 by British troops in Tunis while serving in the Afrika Korps. He was transferred to the United States as a prisoner of war. At the end of the war, Gärtner feared being repatriated to his hometown of Schweidnitz in Lower Silesia, which had by then become a Polish city under communist rule. He decided to escape. On September 22, 1945, he escaped from the prisoner-of-war camp in Deming, New Mexico, by jumping aboard a freight train. This train took him to California. Gärtner then moved from town to town along the West Coast of the United States. He worked as a lumberjack, dishwasher, and laborer. He improved his English, adopted a new identity as Dennis F. Whiles, obtained a Social Security card, and fabricated a biography in which he claimed to have grown up in a Connecticut orphanage after the accidental death of his parents. He eventually settled in Norden, California, where he worked as a ski instructor in the winter and in construction or sales in the summer. It was there that he met Jean Clarke, whom he married in 1964. He adopted her two children from a previous marriage. Throughout this time, the American military continued its search. For forty years, Gärtner remained on the FBI's list of most wanted individuals. However, American authorities considered him to pose no threat, believing he was merely trying to avoid being sent back to Eastern Europe—not out of loyalty to defeated Germany, but out of fear of the new regime. Although Gärtner led a quiet life, his wife grew increasingly worried by his refusal to discuss his past. In 1984, when she threatened to leave him, he confessed the whole truth to her. At her insistence, he decided to make his story public. The following year, he contacted historian Arnold Krammer, a renowned expert on German prisoners of war in the United States. Together, they published a book titled Hitler's Last Soldier in America. Gärtner became famous overnight. He was invited onto several television programs, including The Today Show, where he symbolically "surrendered" to host Bryant Gumbel. The income generated by the book sales ensured him a degree of financial independence. When he came forward, the American government found itself facing a legal vacuum. Gärtner was not an illegal immigrant, since he had been brought to the United States against his will. He had not escaped from a prison in the strict sense, since prisoners of war were, in principle, supposed to be released and repatriated. And because his escape had taken place after the official end of hostilities, it was uncertain whether he would still be considered a prisoner of war in the legal sense. No charges were filed. The FBI stated it no longer had any interest in him, and immigration officials confirmed they had no intention of deporting him. Gärtner was invited to apply for U.S. citizenship. After several years of application, he was naturalized in November 2009. He devoted the end of his life to his passion for painting. Georg Gärtner passed away in 2013 in Loveland, Colorado.
110
756
57
12,903
689,940
Achim Mueller retweeted
17歳で数学界を驚かせた天才少女、今度は一般相対性理論の難問「エーラース=クント予想」を覆す反例を提示 levtech.jp/media/detail_948/ 昨年17歳で[溝畑・竹内予想]を覆したハンナ・カイロさんが今回,一般相対性理論で1962年提唱以来未解決の[測地的完備な重力pp波は必ず平面波]予想に反証し不成立主張
29
696
99
3,726
388,225
Was den Mathematikern derzeit passiert, ist eine Katastrophe. Ihre geistige Arbeit wird schlagartig entwertet, sie werden zu Zuarbeitern der KI. Das könnte das Schicksal der meisten Akademiker in Zukunft sein. Der Respekt vor Denkleistung von Menschen verfällt wie ihr Einkommen
数学界今天彻底破防了,人类最后的智力象牙塔正在坍塌。 陶哲轩刚刚在博客转发了一份炸裂声明:OpenAI 昨天突然毫无预警地向全世界甩出了 700 多份数学论文,声称直接攻克了一大批人类历史上悬而未决的顶级数学难题。 然而,由顶尖学者组成的“人类数学协会”(AHM)不仅没鼓掌,反而集体发起了抵制呼吁。 AHM 声明里的原话极具攻击性: “一次性倾倒 700 篇论文不是学术贡献,这是赤裸裸的权力展示(Demonstration of power)。” 数学家们愤怒的点在于: 1. 学术拒绝服务攻击(DoS): 这相当于企业为了 IPO 造势,把 700 个核弹扔给学界,却让全世界没有薪水的数学家充当免费人肉审稿机。 2. “无法理解的答案”算真理吗? 如果一个世纪难题被黑盒算力+形式化代码暴力破开了,但整个人类社会没有一个人能真正理解其中的直觉与机制——数学到底算不算被推进了? 评论区一句高赞直接把我看愣了: “当年 AlphaGo 碾碎围棋三百年定式的时候,数学家高高在上看热闹;今天轮到自己的象牙塔被推倒了,大家才突然想起来要保卫‘人类的理解’。” 归根结底是数学家的利益受到了侵害。 不过话说回来,当柯尔莫哥洛夫、庞加莱苦思冥想一辈子的直觉,最终被 GPU 阵列的蛮力推导在几个小时内冲刷殆尽;当证明变得像自动化工厂流水线一样廉价,人类对真理的求索究竟还剩什么意义? 你站哪边? A. 赞成 AHM:没有人类理解的证明只是算力垃圾 B. 站 OpenAI:真理不问出处,人类守门人该清醒了
507
91
48
1,029
228,119