the lich
@thelichhh • 2,860 subscribers
data analyst
Videos
8 DGX Sparks on a desk just ran a frontier model at 400K context. A friend who owns the inference budget at a Berlin AI startup, $52K a month in API bills, sent me this at 2 a.m. with one line: "we're buying the hardware." It starts as a disaster. A mountain of cables burying $32K of NVIDIA silicon, nothing talking to anything. 54 minutes later the whole stack answers as 1 machine. 1TB of unified memory. GLM-5.2 loaded across all 8 nodes. The part that made him rewatch twice sits around the 25-minute mark, where 1 networking fix turns 8 boxes into a cluster. Most people quit exactly there. He paused it 4 times. Screenshotted the interface setup. Forwarded it to his CTO without a caption. The cloud bill was never the moat. The cables were.
the lich56,893 просмотров • 3 месяцев назад
A Fields Medalist walked into a UCLA hall and explained prime numbers to people with no mathematics background. Almost nobody watches it. Terence Tao won the Fields Medal, mathematics' highest honour, then a MacArthur Fellowship less than a month later. He was UCLA's first professor to take the Fields. The talk is structure and randomness in the primes. Numbers that look scattered at every scale, yet obey rules strict enough to build cryptography on. That tension is the same one running through every model trained in 2026. Randomness with structure underneath, and the whole question of where one stops and the other starts. Watch how he pitches it. General audience, no prerequisites, one of the strongest living mathematicians choosing clarity over display. A cryptography engineer I know rewatched the middle section twice and said the primes stopped looking arbitrary. Free on YouTube from UCLA, 17 years online. The hardest minds explain most simply. The pattern never changes.
the lich50,394 просмотров • 3 месяцев назад
A Nobel Prize-winning economist was by himself on a dark stage and said to the audience that most of the things they believe about the Great Depression are incorrect. Very few people watch it. Milton Friedman discusses five myths: those concerning the robber barons, the Depression, the public's demand for government, the idea of a free lunch, and who in fact ends up receiving the transferred wealth. He states that the industrialists in the 19th century became wealthy through production rather than through appropriation, and that no other period achieved a faster improvement in the standard of living for the average person. When looking at the transition from the Depression to the free lunch, note that every way a government adopts to generate money eventually comes back to the same person, and he refers to that person. The stage is empty except for a lectern and a microphone, and he is not holding any notes. People repeat the myths and he recorded the arithmetic.
the lich39,177 просмотров • 2 месяцев назад
A Fields Medalist stood on a stage in the 1960s and taught the geometry every gradient moves through. The film went missing for years. This is John Milnor, lecture 1 on differential topology, filmed by the Mathematical Association of America soon after he won the Fields Medal in 1962. He was in his early 30s. The same lectures became Topology from the Differentiable Viewpoint, a book still assigned 60 years later and short enough to read in a week. Smooth maps, manifolds, regular values. The vocabulary under every loss surface, every latent space, every paper with the word manifold in the abstract. Watch what he writes at the top of the board and how little he needs after it. No slides, no notation dumped on the room, each definition doing real work. A research engineer I know worked through the book alongside the film and said the two together beat every modern course he tried. The footage was unavailable for years. It sits free on YouTube now, black and white, 1 fixed camera. The best textbook in the field is a transcript of this room.
the lich29,983 просмотров • 3 месяцев назад
A mathematician solved a problem that stood open for over 100 years, then explained it in an Italian lecture hall to a few dozen people. This is Hong Wang, professor at NYU's Courant Institute and at IHES in Paris, delivering the Colloquium Patavinum on Kakeya sets in 3 dimensions. The question is deceptively small: take a needle of length 1 and rotate it to point in every possible direction. How little space can the set of positions occupy. The conjecture says the answer must fill full dimension. Wang proved it in 3 dimensions. The Salem Prize followed in 2025, the Sadosky Prize in 2026, and an invitation to speak at the International Congress of Mathematicians. Watch how she frames the general statement about unions of tubes. The needle problem falls out as a corollary of something larger. A researcher I know rewatched the setup twice and said it was the clearest account of the result he had found anywhere. Free on YouTube from an Italian mathematics department. The problem was posed in 1917. She closed it, then took questions.
the lich13,485 просмотров • 3 месяцев назад
Richard Hamming stood in front of Navy graduate students in 1995 and explained why smart engineers build the wrong systems. The camera caught 45 minutes of it. Almost nobody watches it. Hamming ran computing at Bell Labs when Bell Labs was inventing the modern world. Error-correcting codes carry his name. Every hard drive, every satellite link, every AI training run leans on his math. This lecture isn't about code. It's about why teams optimize the part and lose the whole. Every startup drowning in AI agents that work in demos and die in production is making the mistake he names here. He was 80 years old. He'd watched 4 decades of brilliant people fail the same way and distilled it into 1 chalk talk for students who mostly forgot it. An engineering director I know stopped the video near the middle, screenshotted the board, and made it his team's wallpaper. The footage sits free on YouTube. Fewer views than a desk lamp review. The systems fail the same way every decade. The lecture doesn't change.
the lich10,846 просмотров • 3 месяцев назад
Больше нет контента для загрузки