Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

Karpathy didn't make a course. He made THE course. 3 hours. Free. Tokenization. Attention. Hallucinations. Tool use. RLHF. DeepSeek. AlphaGo. Every behavior you've ever wondered about in an LLM - where it comes from, why it exists, how it was engineered. The gap between engineers who understand this and... show more

Rahul

129,347 subscribers

630,907 просмотров • 2 месяцев назад •via X (Twitter)

Наука и технологии

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

Andrej Karpathy could have charged $10,000 for this course. He put it on YouTube. The man who built Tesla Autopilot from scratch. Co-founded OpenAI. Understands AI at a level most engineers at Google and Meta never reach. Sat down. Recorded 2 hours. No frameworks. No libraries. No shortcuts. Then dropped it for free. The gap between people who watch it this week and those who save it for later is not 2 hours. It is everything those 2 hours quietly unlock for the rest of your career.

Andrej Karpathy could have charged $10,000 for this course. He put it on YouTube. The man who built Tesla Autopilot from scratch. Co-founded OpenAI. Understands AI at a level most engineers at Google and Meta never reach. Sat down. Recorded 2 hours. No frameworks. No libraries. No shortcuts. Then dropped it for free. The gap between people who watch it this week and those who save it for later is not 2 hours. It is everything those 2 hours quietly unlock for the rest of your career.

NeilXbt

520,162 просмотров • 2 месяцев назад

Andrej Karpathy just built GPT-2 from scratch on camera! Every line. Every component. Every optimization decision. Transformer blocks. Self-attention. MLP. Mixed precision. Flash Attention. AdamW. Distributed training across GPUs. 4 hours. Free. Not a lecture about how GPT-2 was built. The actual build. Running. Training. Evaluated against the original OpenAI checkpoints. The engineers who watch this do not just understand language models. They understand them at the level required to build the next one. The gap between using AI and being capable of building it is not a PhD program. It is 4 hours with the person who makes this look like something any determined engineer can do.

Andrej Karpathy just built GPT-2 from scratch on camera! Every line. Every component. Every optimization decision. Transformer blocks. Self-attention. MLP. Mixed precision. Flash Attention. AdamW. Distributed training across GPUs. 4 hours. Free. Not a lecture about how GPT-2 was built. The actual build. Running. Training. Evaluated against the original OpenAI checkpoints. The engineers who watch this do not just understand language models. They understand them at the level required to build the next one. The gap between using AI and being capable of building it is not a PhD program. It is 4 hours with the person who makes this look like something any determined engineer can do.

NeilXbt

37,720 просмотров • 2 месяцев назад

THE GUY WHO BUILT GIT GAVE A FULL LECTURE AT GOOGLE BECAUSE 99% OF DEVELOPERS USE IT WRONG 70 minutes of pure firepower from Linus Torvalds himself - the man who wrote both Linux and Git from scratch. -> The moment you watch it, you realize why "I'll just figure out merge conflicts on the job" is the dumbest sentence in modern software engineering. He spends the first 10 minutes calling out every other version control system in existence and he's not joking. He literally tells Google engineers using Perforce to find a new tool. Git isn't a "nice-to-have" anymore -> it's the operating system of every codebase you'll ever touch. If the man who built it had to explain it to a room full of Google engineers, you don't get a pass either. Don't forget to bookmark & watch it today.

THE GUY WHO BUILT GIT GAVE A FULL LECTURE AT GOOGLE BECAUSE 99% OF DEVELOPERS USE IT WRONG 70 minutes of pure firepower from Linus Torvalds himself - the man who wrote both Linux and Git from scratch. -> The moment you watch it, you realize why "I'll just figure out merge conflicts on the job" is the dumbest sentence in modern software engineering. He spends the first 10 minutes calling out every other version control system in existence and he's not joking. He literally tells Google engineers using Perforce to find a new tool. Git isn't a "nice-to-have" anymore -> it's the operating system of every codebase you'll ever touch. If the man who built it had to explain it to a room full of Google engineers, you don't get a pass either. Don't forget to bookmark & watch it today.

slash1s

214,079 просмотров • 1 месяц назад

Anthropic pays $750,000 a year for engineers who can build LLMs from scratch. Not prompt engineers. Not fine-tuners. Not RAG specialists. Engineers who understand the architecture well enough to build the thing from zero. Stanford just put a 2-hour lecture on exactly that on the internet for free. Scaling laws. Data collection. Architecture design. Post-training alignment. The theoretical foundation behind every model you use daily taught by the people who research it at one of the best CS programs on earth. Free. From Stanford. Two hours. The engineers earning $750,000 at Anthropic understand everything in this lecture. Most people using AI every day understand none of it. Watch it tonight. Full implementation guide in the article below.

Anthropic pays $750,000 a year for engineers who can build LLMs from scratch. Not prompt engineers. Not fine-tuners. Not RAG specialists. Engineers who understand the architecture well enough to build the thing from zero. Stanford just put a 2-hour lecture on exactly that on the internet for free. Scaling laws. Data collection. Architecture design. Post-training alignment. The theoretical foundation behind every model you use daily taught by the people who research it at one of the best CS programs on earth. Free. From Stanford. Two hours. The engineers earning $750,000 at Anthropic understand everything in this lecture. Most people using AI every day understand none of it. Watch it tonight. Full implementation guide in the article below.

CyrilXBT

36,692 просмотров • 1 месяц назад

🚨 OpenAI 's own engineers just showed how to actually use OpenAI Codex properly. 60 minutes. free. built by the people who contribute to made it. watch the masterclass. bookmark it. worth more than every $900 coding course you almost bought. you’ve been using Codex like a simple coding tool… while it’s actually a full software engineering system. watch this, it could the best 62 minutes of your life:

🚨 OpenAI 's own engineers just showed how to actually use OpenAI Codex properly. 60 minutes. free. built by the people who contribute to made it. watch the masterclass. bookmark it. worth more than every $900 coding course you almost bought. you’ve been using Codex like a simple coding tool… while it’s actually a full software engineering system. watch this, it could the best 62 minutes of your life:

ZARA

296,432 просмотров • 2 месяцев назад

MIT HANDED ITS DEEP LEARNING COURSE TO A FRONTIER-LAB ENGINEER FOR 68 MINUTES BECAUSE 90% OF PEOPLE SHIPPING AI CODE CAN'T EXPLAIN HOW THE MODEL ACTUALLY WORKS This is Maxime Labonne. He runs post-training at Liquid AI and wrote the LLM Engineer's Handbook. MIT gave him the room to break down the engine sitting inside every coding agent you've ever prompted. Twenty minutes in it stops being abstract. You finally see why the model confidently invents things that don't exist, why context is everything, and why "Just tell it to try again" sometimes fixes it and sometimes makes it worse. In 2026 "I use AI to code" stopped being a skill. Knowing why the model behaves the way it does -> tokens, context, post-training, where it quietly breaks -> is what separates someone who ships from someone babysitting a black box. Understanding an LLM isn't a research-team luxury anymore -> it's the difference between driving the agent and being driven by it. Anyone can prompt. The person who knows what's under the hood is the one still standing when the prompt stops working. Save this one & actually finish it ↓

MIT HANDED ITS DEEP LEARNING COURSE TO A FRONTIER-LAB ENGINEER FOR 68 MINUTES BECAUSE 90% OF PEOPLE SHIPPING AI CODE CAN'T EXPLAIN HOW THE MODEL ACTUALLY WORKS This is Maxime Labonne. He runs post-training at Liquid AI and wrote the LLM Engineer's Handbook. MIT gave him the room to break down the engine sitting inside every coding agent you've ever prompted. Twenty minutes in it stops being abstract. You finally see why the model confidently invents things that don't exist, why context is everything, and why "Just tell it to try again" sometimes fixes it and sometimes makes it worse. In 2026 "I use AI to code" stopped being a skill. Knowing why the model behaves the way it does -> tokens, context, post-training, where it quietly breaks -> is what separates someone who ships from someone babysitting a black box. Understanding an LLM isn't a research-team luxury anymore -> it's the difference between driving the agent and being driven by it. Anyone can prompt. The person who knows what's under the hood is the one still standing when the prompt stops working. Save this one & actually finish it ↓

slash1s

236,758 просмотров • 1 месяц назад

New short course: Attention in Transformers: Concepts and Code in PyTorch. Last week we released a course on how LLM transformers work. This week, go deeper and learn about the technical ideas behind the attention mechanism, and see how to code it in PyTorch. This course is built with Joshua Starmer, Founder and CEO of StatQuest. The attention mechanism was a breakthrough that led to transformers, the architecture powering large language models like ChatGPT. Transformers, introduced in the 2017 paper: "Attention is All You Need" by Viswani and others, took off because of its highly scalable design. In this course, you’ll learn how the attention mechanism, a key element of transformer-based LLMs, works and implement it in PyTorch. You'll develop deep intuition about building reliable, functional, and scalable AI applications. What you will do: - Understand the evolution of the attention mechanism, a key breakthrough that led to transformers. - Learn the relationships between word embeddings, positional embeddings, and attention. - Learn about the Query, Key, and Value matrices, and how to produce and use them in attention. - Walk through the math required to calculate self-attention and masked self-attention to learn why and how they work. - Understand the difference between self-attention and masked self-attention and how one is used in the encoder to build context-aware embeddings and the other is used in the decoder for generative outputs. - Learn the details of the encoder-decoder architecture, cross-attention, and multi-head attention and how they are all incorporated into a transformer. - Use PyTorch to code a class that implements self-attention, masked self-attention, and multi-head attention. There're lots of exciting technical details in this course. Please sign up here:

New short course: Attention in Transformers: Concepts and Code in PyTorch. Last week we released a course on how LLM transformers work. This week, go deeper and learn about the technical ideas behind the attention mechanism, and see how to code it in PyTorch. This course is built with Joshua Starmer, Founder and CEO of StatQuest. The attention mechanism was a breakthrough that led to transformers, the architecture powering large language models like ChatGPT. Transformers, introduced in the 2017 paper: "Attention is All You Need" by Viswani and others, took off because of its highly scalable design. In this course, you’ll learn how the attention mechanism, a key element of transformer-based LLMs, works and implement it in PyTorch. You'll develop deep intuition about building reliable, functional, and scalable AI applications. What you will do: - Understand the evolution of the attention mechanism, a key breakthrough that led to transformers. - Learn the relationships between word embeddings, positional embeddings, and attention. - Learn about the Query, Key, and Value matrices, and how to produce and use them in attention. - Walk through the math required to calculate self-attention and masked self-attention to learn why and how they work. - Understand the difference between self-attention and masked self-attention and how one is used in the encoder to build context-aware embeddings and the other is used in the decoder for generative outputs. - Learn the details of the encoder-decoder architecture, cross-attention, and multi-head attention and how they are all incorporated into a transformer. - Use PyTorch to code a class that implements self-attention, masked self-attention, and multi-head attention. There're lots of exciting technical details in this course. Please sign up here:

Andrew Ng

132,220 просмотров • 1 год назад

When this video was first texted to me, I assumed it was from Iran. The way she screams, she was wearing a hijab, and the way the law enforcement approached her it looked no different from the way Islamic Republic authorities arrest women. It's from Massachusetts, of course. There's no reason to detain someone in this manner who isn't a danger to the public. None at all.

When this video was first texted to me, I assumed it was from Iran. The way she screams, she was wearing a hijab, and the way the law enforcement approached her it looked no different from the way Islamic Republic authorities arrest women. It's from Massachusetts, of course. There's no reason to detain someone in this manner who isn't a danger to the public. None at all.

Yashar Ali 🐘

327,617 просмотров • 1 год назад

New short course on Reinforcement Learning from Human Feedback! RLHF is one of the key techniques that led to the rise of modern LLMs. It is used to align LLMs with human preferences, to make them more honest, helpful and harmless, by (i) learning a reward function that mimics human preferences, as expressed in human-provided labels, then, (ii) tuning an LLM to generate outputs that receive a high reward. In this course, taught by Nikita Namjoshi, Developer Advocate for GenAI at Google Cloud, you'll learn the details of how RLHF works, including how to apply it to tune an LLM for your own applications. You'll also use an open source library to tune a base LLM to align with human preferences expressed in a training set, and evaluate the tuned model by comparing its responses before and after RLHF-tuning. Please sign up here!

New short course on Reinforcement Learning from Human Feedback! RLHF is one of the key techniques that led to the rise of modern LLMs. It is used to align LLMs with human preferences, to make them more honest, helpful and harmless, by (i) learning a reward function that mimics human preferences, as expressed in human-provided labels, then, (ii) tuning an LLM to generate outputs that receive a high reward. In this course, taught by Nikita Namjoshi, Developer Advocate for GenAI at Google Cloud, you'll learn the details of how RLHF works, including how to apply it to tune an LLM for your own applications. You'll also use an open source library to tune a base LLM to align with human preferences expressed in a training set, and evaluate the tuned model by comparing its responses before and after RLHF-tuning. Please sign up here!

Andrew Ng

205,537 просмотров • 2 лет назад

IN 1999 MIT FILMED A MATH LECTURE THAT QUIETLY BECAME THE FOUNDATION OF EVERY AI MODEL YOU'VE EVER USED AND ALMOST NO ONE WAS TAUGHT TO SEE IT THAT WAY 39 minutes from Gilbert Strang, who taught this at MIT for over 60 years -- the linear algebra course an entire generation of engineers and data scientists grew up on. -> The shift it creates: you stop seeing matrices as boring grids of numbers and start seeing them as the language of space, data, and motion itself. School drilled you to crunch matrices by hand and never told you why. Strang shows you what they actually mean. Every neural net, every embedding, every model you prompt is linear algebra running underneath. The math you skipped is the engine of the thing you use all day. Memorizing the steps was never the skill -> seeing what the numbers do is. This is where it finally clicks. Most people fear linear algebra and move on. The ones who watched this see straight into how AI actually works. Bookmark & Watch it today, this one's a legend ↓

IN 1999 MIT FILMED A MATH LECTURE THAT QUIETLY BECAME THE FOUNDATION OF EVERY AI MODEL YOU'VE EVER USED AND ALMOST NO ONE WAS TAUGHT TO SEE IT THAT WAY 39 minutes from Gilbert Strang, who taught this at MIT for over 60 years -- the linear algebra course an entire generation of engineers and data scientists grew up on. -> The shift it creates: you stop seeing matrices as boring grids of numbers and start seeing them as the language of space, data, and motion itself. School drilled you to crunch matrices by hand and never told you why. Strang shows you what they actually mean. Every neural net, every embedding, every model you prompt is linear algebra running underneath. The math you skipped is the engine of the thing you use all day. Memorizing the steps was never the skill -> seeing what the numbers do is. This is where it finally clicks. Most people fear linear algebra and move on. The ones who watched this see straight into how AI actually works. Bookmark & Watch it today, this one's a legend ↓

slash1s

431,502 просмотров • 29 дней назад

Gun Hill Road in the Bronx. He had tried to board for free, and was - let's say - ejected. Aside from the basic principle that nothing in life is "free", and the ones who should pay for the service are the ones who use it, this is why you don't want to make bus rides free.

Gun Hill Road in the Bronx. He had tried to board for free, and was - let's say - ejected. Aside from the basic principle that nothing in life is "free", and the ones who should pay for the service are the ones who use it, this is why you don't want to make bus rides free.

Crime In NYC

60,905 просмотров • 9 месяцев назад

I've seen a lot of people who don't understand the Armalite AR17. It's a rare gun, with few existing videos of it shooting. So I made a video about it. It's over an hour long, and should help the gun make more sense. I'll link the video below this post.

I've seen a lot of people who don't understand the Armalite AR17. It's a rare gun, with few existing videos of it shooting. So I made a video about it. It's over an hour long, and should help the gun make more sense. I'll link the video below this post.

Navi of Boomhandia

12,515 просмотров • 1 год назад

AN MIT RESEARCHER PROVED GIT ISN'T HARD BECAUSE YOU'RE BAD AT IT -- IT'S HARD BECAUSE IT WAS DESIGNED THAT WAY 27 minutes from a PhD researcher in MIT's Software Design Group, using actual design theory to show why the tool that confuses everyone confuses everyone for a reason. -> The moment it lands, years of feeling stupid evaporate. The gap between what git's commands say and what they actually do was never in your head. It's baked into the tool. He maps the difference between what you think a command does and what git really does underneath. Once you see that gap, the confusion finally has a name and it stops being yours to carry. Struggling with git was never a skills issue -> it's a design issue, and knowing where the model lies to you is what turns panic into control. And as AI agents fire off commits and rebases you didn't write, the person who understands where git misleads is the one who untangles the mess. You were never bad at git. You were just never shown where it was built to trip you. Bookmark & Watch it today ↓

AN MIT RESEARCHER PROVED GIT ISN'T HARD BECAUSE YOU'RE BAD AT IT -- IT'S HARD BECAUSE IT WAS DESIGNED THAT WAY 27 minutes from a PhD researcher in MIT's Software Design Group, using actual design theory to show why the tool that confuses everyone confuses everyone for a reason. -> The moment it lands, years of feeling stupid evaporate. The gap between what git's commands say and what they actually do was never in your head. It's baked into the tool. He maps the difference between what you think a command does and what git really does underneath. Once you see that gap, the confusion finally has a name and it stops being yours to carry. Struggling with git was never a skills issue -> it's a design issue, and knowing where the model lies to you is what turns panic into control. And as AI agents fire off commits and rebases you didn't write, the person who understands where git misleads is the one who untangles the mess. You were never bad at git. You were just never shown where it was built to trip you. Bookmark & Watch it today ↓

slash1s

138,043 просмотров • 1 месяц назад

$PLTR Dan Ives on why retail was able to understand Palantir before institutional investors: "People didn't understand the culture of Palantir. We're talking about the elite of the elite. From top down to engineers to the merch and everything else. Institutional investors were stuck in group-think when it came to Palantir as an investment in 2022. It wasn't part of the cool kids club. It's very easy to miss something if you get caught up in the consensus view."

$PLTR Dan Ives on why retail was able to understand Palantir before institutional investors: "People didn't understand the culture of Palantir. We're talking about the elite of the elite. From top down to engineers to the merch and everything else. Institutional investors were stuck in group-think when it came to Palantir as an investment in 2022. It wasn't part of the cool kids club. It's very easy to miss something if you get caught up in the consensus view."

amit

124,569 просмотров • 1 год назад

Bravo to Marine Le Pen for blasting Islamic veil in France: “I consider the Islamist veil to be a marker of this totalitarian ideology. I thus wish to free every women on our national territory from it. I said it this morning to neighbourhood under the pressure from Islamists (to wear veil) is that of course on the women who don't wear it, isolating the women who don't wear one. And from that point It's a great loss of freedom for women, which is in fact freedom of choice."

Bravo to Marine Le Pen for blasting Islamic veil in France: “I consider the Islamist veil to be a marker of this totalitarian ideology. I thus wish to free every women on our national territory from it. I said it this morning to neighbourhood under the pressure from Islamists (to wear veil) is that of course on the women who don't wear it, isolating the women who don't wear one. And from that point It's a great loss of freedom for women, which is in fact freedom of choice."

Imtiaz Mahmood

506,924 просмотров • 2 лет назад

So here is a video of her doing a paid partnership with a brand and yes of course mentions her "partner" and how she deals with anxiety and grief. Again making it all about herself. And she was the same person who said anxiety was made up in her head. If you cant see how she will say anything to make $$$ and use liam to do it, then open your eyes. This is the same woman who never once used her platform to talk about liam struggles while he was alive.

Jade²⁸❯❯❯❯

10,485 просмотров • 1 месяц назад

most people spend hours describing what they want to Claude Code this guy just browsed a catalog of thousands of animated UI components, found one he liked, and hit "copy prompt" that's it. the prompt goes into Claude Code and it builds it. > 21st(.)dev - a library of community-made components built for React, Next.js, Tailwind > every component has a ready-made prompt attached > you pick the effect, copy, paste, done no describing animations from scratch. no explaining what "smooth scroll" or "hover glow" means. someone already figured out the exact prompt that produces it. gap between people who know this exists and people who don't is measured in hours per project full walkthrough in the article below

most people spend hours describing what they want to Claude Code this guy just browsed a catalog of thousands of animated UI components, found one he liked, and hit "copy prompt" that's it. the prompt goes into Claude Code and it builds it. > 21st(.)dev - a library of community-made components built for React, Next.js, Tailwind > every component has a ready-made prompt attached > you pick the effect, copy, paste, done no describing animations from scratch. no explaining what "smooth scroll" or "hover glow" means. someone already figured out the exact prompt that produces it. gap between people who know this exists and people who don't is measured in hours per project full walkthrough in the article below

monokern

64,484 просмотров • 11 дней назад

Mad Men creator said the show was never about advertising. It was about the gap between what people want and what they can actually have. "Happiness is always in that gap in between," he said. Don is a man who has everything the era told him to want, and cannot feel any of it.

Mad Men creator said the show was never about advertising. It was about the gap between what people want and what they can actually have. "Happiness is always in that gap in between," he said. Don is a man who has everything the era told him to want, and cannot feel any of it.

Emir Han

679,996 просмотров • 1 месяц назад

Anthropic Managed Agents Lead: "Over 90% of Anthropic engineers already build with self-improving loops. My agentic loops run for hours and cost almost nothing." In 40 minutes she builds an AI agent from scratch and shows how engineers at Anthropic use them daily. This video is worth more than any $600 course on agents. Watch it today, then read the guide on building loops below.

Anthropic Managed Agents Lead: "Over 90% of Anthropic engineers already build with self-improving loops. My agentic loops run for hours and cost almost nothing." In 40 minutes she builds an AI agent from scratch and shows how engineers at Anthropic use them daily. This video is worth more than any $600 course on agents. Watch it today, then read the guide on building loops below.

Anatoli Kopadze

151,932 просмотров • 14 дней назад

Walter Isaacson reveals how Elon Musk fired 85% of Twitter's employees in three rounds and why "He brought in the engineers and figured the amount of people doing Tesla full self-driving and autopilot software was about one-tenth of what was doing software for Twitter. He said this can't be the case" "He fired 85% in three different rounds. First was firing people because they had a team from Tesla's autopilot team grading the code written in the past year. Then he fired people who didn't seem totally all in or loyal. Then another round of layoffs" "At each step, almost everybody says that's enough, it's going to destroy things. They were partly right. There was degradation of the service, but not as much as half the services I use" "He has an algorithm. Question every requirement, then delete delete delete. And the corollary, if you don't end up adding back 20% of what you deleted, you didn't delete enough in the first round" "You asked me if he overdid it. He probably overdid it by 20%, which is his formula"

Walter Isaacson reveals how Elon Musk fired 85% of Twitter's employees in three rounds and why "He brought in the engineers and figured the amount of people doing Tesla full self-driving and autopilot software was about one-tenth of what was doing software for Twitter. He said this can't be the case" "He fired 85% in three different rounds. First was firing people because they had a team from Tesla's autopilot team grading the code written in the past year. Then he fired people who didn't seem totally all in or loyal. Then another round of layoffs" "At each step, almost everybody says that's enough, it's going to destroy things. They were partly right. There was degradation of the service, but not as much as half the services I use" "He has an algorithm. Question every requirement, then delete delete delete. And the corollary, if you don't end up adding back 20% of what you deleted, you didn't delete enough in the first round" "You asked me if he overdid it. He probably overdid it by 20%, which is his formula"

Jaynit

748,307 просмотров • 1 месяц назад