bro created an entire 16-hour free youtube playlist on how to build a DeepSeek model from scratch. it goes over the papers, explains the theory, and implements the code. Syllabus: → attention mechanism fully explained → multi-head latent attention → grouped query attention → everything about positional encodings → mixture of experts (MoE) just start today with a laptop and motivation. playlist: piped.video/playlist?list=PL…

My thesis has always been that screens trick you into thinking you're productive while sitting. If you make a rule that you can only scroll while standing, your phone addiction will die immediately, violently even, with no other changes necessary.

Ever since I discovered that Harvard University was partly built on money made from the opium trade in China, I began to understand geopolitics at a deeper level. And it wasn’t just Harvard, Yale too. The founder of the Skull and Bones society, one of the most secretive and influential networks in American history, was himself from an opium dealing family. In fact, parts of the early American railway system were financed with profits from that same trade, selling narcotics to a nation that had once been the world’s industrial and cultural powerhouse. You can draw a near-straight line from America’s early economic miracle to the opium wealth extracted from China. Yet, the men who orchestrated this weren’t called drug dealers, they were called merchants, financiers, and philanthropists. They laundered blood money into legitimacy, endowing universities, libraries, and churches, and in return, received moral prestige. China, once the richest and most advanced civilization on Earth, was reduced to poverty and humiliation because of that trade. Historical context is everything if you want to understand world affairs. The same West that crippled China with opium now lectures it over fentanyl. It's really a full circle moment. It's only in Africa that we smile at our historical saboteurs. Only in Africa do we entertain the same powers that impoverished us, letting them dictate how we should think, trade, and govern. Even their so-called occasional concern for us drips with paternalism. For me, I admire nations that remember their humiliation and have the strategic patience to sometimes wait for generations to avenge it. That’s the kind of historical memory and disciplined vengeance I wish for future Africans. I’ve already lost faith in the current crop, but perhaps the next generation could make it happen.

Oil is the most Lovecraftian thing that actually exists. You're telling me that there's a black ichor under the earth, made from the ancient dead, whose burning can realize all the dreams of man but only at the price of slowly returning the earth to its primordial state?

This is the most brilliant idea I've heard in maybe 20 years. It is genius. Between Iranian architecture and Chinese philosophy of knowledge, the West looks more and more uncivilised and anachronistic

I never felt like I belonged to this era until AI came around

here is a mix

Muchos citan a Marx para decir que la automatización “disolverá la ley del valor”. Pero Marx no predijo que las máquinas destruirían el capitalismo por sí solas. Lo que señaló fue algo más profundo: > el capital cava su propia tumba al intentar prescindir del trabajo humano.

La mal llamada “ciencia de datos” no es una ciencia en sentido estricto. No tiene un objeto de estudio propio ni desarrolla teorías explicativas verificables; más bien, aplica herramientas provenientes de la estadística, la computación y las matemáticas para analizar información y resolver problemas prácticos. Su propósito es instrumental, no epistémico: busca utilidad, no verdad. Por tanto, es más correcto considerarla una disciplina tecnológica interdisciplinaria, al mismo nivel que la ingeniería de software o la actuaría, que aplican ciencias, pero no constituyen una por sí mismas. El nombre “ciencia de datos” es atractivo y moderno, pero epistemológicamente incorrecto: lo que hace no es generar conocimiento científico, sino usar el conocimiento científico existente para extraer valor de los datos.

you guys should actually just go run the code. It's literally just a pip install. Install it and train a model on your computer in 60 seconds. Then literally just go read the code. It's actually simple Lots of AI salesmen selling complicated bullshit. This is simple and good

You only need 20 videos

La Semilla del Deseo | Cortometraje Emocional de Duelo y Aceptación piped.video/bYqUjpNSa78?si=dlej… via @YouTube

We’re not meant to live like this bro we are not

🚨 Brutally honest truths about YouTube Automation NOBODY talks about (bookmark for later) If you're doing faceless YouTube in 2025, this thread is going to save you months of pain 🧵

Byung-Chul: La universidad es una empresa neoliberal que todo lo sacrifica en el altar del desempeño. Reparte puntos. Los catedráticos son vendedores, los estudiantes son clientes que evalúan a los profesores. Este es el colapso de la cultura. elmundo.es/cultura/literatur…

it's that simple

🚨 Perplexity just dropped something wild. It’s called “Perplexity at Work” their official guide to actually getting more done with AI. Not another “productivity tips” doc this is the real framework their own teams use to: → Block distractions & reclaim focus → Scale yourself like a 5-person team → Turn AI from noise into results It’s clean, practical, and honestly the most useful thing I’ve read on using AI for work not just chat prompts. Comment “Guide” and I’ll send you the full PDF (it’s 100% free from Perplexity)