Michel Yamagishi

Over 32,000 medieval manuscripts transcribed in four months using AI medievalists.net/2026/01/32000-… #medievalmanuscripts

Can AI develop true reasoning skills? (A pure mathematician’s perspective) youtube.com/watch?v=5k0cF6…

As médias dos cursos de Medicina no ENAMED não foram detalhadas na cobertura jornalística sobre o exame. No entanto, esses dados fazem parte das informações oficiais divulgadas e podem ser consultados de forma organizada: O site enamed.info reúne as médias por curso, permitindo uma leitura mais cuidadosa dos resultados e contribuindo para uma análise mais qualificada da avaliação da formação médica no Brasil. Para gestores educacionais, docentes, pesquisadores, estudantes e demais interessados em políticas de avaliação do ensino superior, trata-se de uma fonte útil e de fácil acesso. 👉 enamed.info

Recentemente descritos pela ciência, os misteriosos cogumelos são encontrados em diferentes partes do mundo, mas provocam nas pessoas exatamente as mesmas visões: bbc.in/3NDiHex

@QuantaMagazine @nattyover We don't need new physics, but a trustworthy Philosophy.

In the first of our new series of curiosity-driven essays, Qualia, @nattyover asks particle physicists whether the field is facing a profound crisis. quantamagazine.org/is-particle-ph…

@DavidSHolz or worse

all of evolution implies - we can become better

@jdlichtman Silence and solitude are the gateways to deep understanding.

🍎 Isaac Newton discovered calculus while in lockdown during the Black Plague 🪶Andre Weil proved the Riemann Hypothesis for curves while in prison during WWII There's a lesson in there somewhere

@TheCinesthetic Amadeus

What movie has the perfect soundtrack?

“What was you favorite equation from 2025?” Me:

@mathematics_inc I hope it will contribute to resolving long-standing open problems, including the twin prime conjecture and the Goldbach conjecture.

🎀 Terence Tao is partnering with Math, Inc. 🎀 as the inaugural Veritas Fellow — to formalize estimates in number theory. In analytic number theory, the literature contains a large web of explicit estimates. But that web is not immediately interoperable. In practice, results come in three layers: Primary estimates: These are foundational inputs such as zero-free regions for the Riemann zeta function. They often depend on substantial computation and careful numerical optimization. Secondary estimates: Many papers take a primary input (e.g., a zero-free region) and convert it into reusable consequences, such as counting primes in short intervals. These become core building blocks used throughout the subject. Tertiary estimates: Further work then applies those secondary building blocks to frontier number-theoretic problems, e.g. representing integers as sums of three primes. The difficulty is that these layers do not update cleanly over time. A tertiary paper may rely on the best primary estimate available at the time. But years later improved computations refine the primary input, without being systematically propagated through the secondary and tertiary chain. As a result, the “same theorem with updated constants” is often unknown. The goal is to formalize key papers across these layers and then abstract them so their dependencies become explicit, composable, and machine-checkable. The long-term vision is to create a living network of implications: when a primary estimate improves, every downstream implication is automatically upgraded. This will transform the mathematical literature into modular software. Number theory is a strong test case because its estimates has a relatively clear structure, and a shared set of standard inputs and outputs. But in many areas such as PDEs, researchers constantly spend effort on modification: adapting lemmas and hypotheses, translating between incompatible frameworks, “fitting square pegs into round holes.” A composable, machine-verified implication network directly targets this friction. The same infrastructure is poised to scale to other fields and enable crowdsourced, large-scale projects that are currently hard to coordinate. A classic example is the classification of finite simple groups: a decades-long effort distributed across many contributors, with inevitable complexity around bookkeeping, integration, and confidence in completeness. With modern tooling, we envision tackling moonshots of comparable scope: many contributors handling diverse cases, and automated systems gluing the pieces together. The field becomes a live progress dashboard that records what is proved, what remains, and exactly which dependencies each component requires. This opens up the possibility for a much faster-pace and engaging way to do mathematics. Watch Tao's outline on YouTube: youtube.com/watch?v=KeX9Jq… math.inc/a-conversation…

@jdlichtman Sometimes, reframing the problem opens up entirely new paths.

"There is some speculation as to some possible ways to solve the conjecture, but nothing is really promising." (5/5) youtube.com/watch?v=MXJ-zp…

The Riemann Hypothesis (RH) has changed in perceived difficulty over time: In 1900, Hilbert thought RH would be one of the easier problems to solve on his list. Today, RH is viewed difficult because there aren't any promising proof strategies. 🧵 (1/5)

@Ganeshuor By the way, this is not intended as a criticism of specialists, but as an observation about a broader cultural shift that affects everyone without exception.

If Ramanujan’s letter were sent today, it would likely be filtered straight into the spam folder. And even if it somehow reached the inbox, it would probably remain unread—few people now have the time or willingness to carefully engage with the work of others, especially when it comes from someone entirely unknown in the field.

This is the letter that transformed Ramanujan’s life—a bold, unsolicited message he sent to G. H. Hardy in 1913.