StarZero All shows Try the podcast clipper →

Andrej Karpathy

Founder, Eureka Labs · 0 appearances on the record.

computed by AI from the episodes · how this works → · full disclaimer →

scientistengineerfounderexecutiveacademic@karpathy ↗karpathy.ai ↗Wikipedia ↗YouTube ↗

Andrej Karpathy was a founding member of OpenAI and the former Senior Director of AI at Tesla, where he led the computer vision team for Autopilot. He also founded the AI-native education platform Eureka Labs and is known for creating open-source AI education projects like nanoGPT.

14statements → 6claims → 0claims resolved → —average certainty → —average debate potential → 115said about them ↓

1 not checkable as stated how the 6 claims stand · each chip opens the sources

1 prediction · 5 assertions · 2 opinions · 1 insight · 5 disclosures · every statement was checked. The prediction and assertions are the 6 claims: statements the public record can support or contradict. 5 are resolved, and 1 names no date, number or outcome precise enough to check. Everything else (opinions, insights, what ifs, disclosures) can never be settled by the record, so it carries no assessment.

The record, in short

What the tape says about how Andrej argues and how the claims held up. Everything they said, and everything said about them, is in the tabs below.

Their most notable supported claim

Assertion Supported
Karpathy: llm.c trains GPT-2 on one H100 node in 24 hours for $600
“You can train it on a single node of H-one-hundreds in about 24 hours, and that costs roughly 600 dollars.”
Andrej Karpathy Sep 21, 2024 ▶ 18:02 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE

Expressed certainty vs assessment result

none yet certainty 1
none yet certainty 2
100% certainty 3
100% certainty 4
100% certainty 5

weighted support: a fully supported claim counts one, a partly supported claim counts half. Each filled bar is clickable and opens exactly those claims; "none yet" means nothing said at that certainty level has resolved yet

Everything Andrej Karpathy said on Latent Space that made the record, most notable first. Filter by type, assessment or year in the ledger →

Prediction Not checkable as stated
Karpathy predicts LLMs will act as compilers generating bare-metal CUDA code
“If LLINs are about to become much better at coding over time, then I think you can expect that the LLIN could actually do this for any custom application over time. And so the LLINs could act as a kind of compiler What you're interested in, they're gonna do al…”
Andrej Karpathy Sep 21, 2024 ▶ 21:54 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Insight
Karpathy: Python and PyTorch are crutches for finite human intelligence
“The use of Python and PyTorch and everything else is just a crutch, because we humans are finite. We have finite knowledge, intelligence, and attention.”
Andrej Karpathy Sep 21, 2024 ▶ 22:17 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Assertion Supported
Karpathy: llm.c trains GPT-2 on one H100 node in 24 hours for $600
“You can train it on a single node of H-one-hundreds in about 24 hours, and that costs roughly 600 dollars.”
Andrej Karpathy Sep 21, 2024 ▶ 18:02 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Assertion Supported
Karpathy: llm.c was 20% faster and used 30% less memory than PyTorch
“At the time of that post, we were using, in LL and that's in 30% less memory, and we were 20% faster in training, just the truth.”
Andrej Karpathy Sep 21, 2024 ▶ 19:10 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Disclosure
Karpathy: llm.c avoids tensor abstractions in favor of raw float arrays
“And one thing I'd really like to do in LN.c is I just want to keep things simple. I don't want to create a tensor abstraction. I don't want to create any abstraction, really. It's just float arrays and operations on float arrays.”
Andrej Karpathy Sep 21, 2024 ▶ 6:58 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Opinion
Karpathy: The popular PMPP textbook lacks advanced CUDA optimization techniques
“PMPP is actually quite good but also, I think, still kind of like mostly on the beginner level, because a lot of the CUDA code that we ended up developing in the lifetime of the LMC project, you would not find those things in, in this book, actually.”
Andrej Karpathy Sep 21, 2024 ▶ 12:13 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Disclosure
Karpathy: llm.c rejects performance PRs if added complexity harms developer onboarding
“A lot of lmc just kind of, like balancing the improvement and speed with the complexity of what you're actually introducing, and so I've actually rejected a lot of PRs because of that, because the code starts to get crazy, and I think that decreased the amount…”
Andrej Karpathy Sep 21, 2024 ▶ 17:25 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Disclosure
Karpathy: llm.c trains transformers in C with minimal C++
“We're training transformers in C at a pinch of C++.”
Andrej Karpathy Sep 21, 2024 ▶ 1:38 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Opinion
Karpathy: Without PyTorch, developers are 'naked in the world'
“So, PyTorch is really, really nice, and this is just some of the things that PyTorch offers. So, without PyTorch, we're kind of naked in the world, right?”
Andrej Karpathy Sep 21, 2024 ▶ 5:09 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Assertion Supported
Karpathy: llm.c avoids runtime crashes by pre-allocating memory statically without dependencies
“It's a single file C. There's no dependencies whatsoever. It compiles instantly. It runs instantly. All the memory is just allocated in a single blob. So if you start stepping, there's no way you're gonna boom later. It's all preplanned. It's fully determinist…”
Andrej Karpathy Sep 21, 2024 ▶ 9:17 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Disclosure
Karpathy: llm.c was created while jet-lagged on vacation in the Maldives
“All of this work that I described so far, I was on vacation and while I was jet-lagged in Maldives. So I, basically it's perfect because you wake up at one a.m., and there's nothing to do. So you write stuff like LLN.C.”
Andrej Karpathy Sep 21, 2024 ▶ 10:03 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Assertion Supported
Karpathy: llm.c achieves nearly 50% MFU on single-node GPT-2 training
“We have almost a 50% NFU here on one node, which is quite good.”
Andrej Karpathy Sep 21, 2024 ▶ 18:45 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Disclosure
Karpathy: llm.c will support LLaMA 3.1 training very soon
“We will have Lama 3.1 training in Lama.c very, very soon.”
Andrej Karpathy Sep 21, 2024 ▶ 20:28 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE
Assertion Supported
Karpathy: llm.c is only about 3,000 lines of code
“It's only maybe like, I think, 3000 lines of code, basically C mostly.”
Andrej Karpathy Sep 21, 2024 ▶ 21:11 llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE

The other half of the tape: Andrej Karpathy's own voice is left out of every number here. Other people bring the name up 110 times in 51 episodes on Latent Space. 5 statements on the record name them. every mention, with the transcript →

Who brings them up most Shawn Wang 18Mikhail Parakhin 6Alessio Fanelli 5Raiza Martin 4Lance Martin 4Gavriel Cohen 4Jeremy Howard 3Eugene Cheah 3

Statements about Andrej Karpathy, by other people (5)

Assertion Not checkable as stated
Parakhin: Shopify built agent memory systems before Karpathy proposed them
“Just even yesterday, Andrej Karpathy was famous in tweeting about, oh, there's some ways that you can organize your agents to store the data and then look up the data so that you don't have to research or lose context every time. And a little bit tongue-in-che…”
Mikhail Parakhin Apr 22, 2026 ▶ 3:48 AI-Native Engineering: 100% adoption, 5x search throughput, unlimited tokens — Mikhail Parakhin
Insight
Doshi: Learn AI via Andrej Karpathy's YouTube tutorials over books
“You shouldn't take a book. You should just go to YouTube and visit Kaparthi's class. And just do it, grind through it.”
Suhail Doshi Jan 2, 2024 ▶ 1:07:33 The AI-First Graphics Editor - with Suhail Doshi of Playground AI
Assertion Not checkable as stated
Howard: Tesla and OpenAI Scholars used Fast.ai courses for deep learning training
“Andre Capathy grabbed me when I saw him at NeurIpes a few years ago, and he's like, I have to tell you, thanks to the fast AI courses, when people come to Tesla, and they need to know more about deep learning, we always send them to your course. And the OpenAI…”
Jeremy Howard Oct 20, 2023 ▶ 28:15 The End of Finetuning — with Jeremy Howard of Fast.ai
Assertion Not checkable as stated
Swyx: Andrej Karpathy has promised to appear on Latent Space
“Is a supporter of ours that is, that has promised to be on at some point Andre.”
Shawn Wang Dec 12, 2025 ▶ 1:08:43 AI to AE's: Grit, Glean, and Kleiner Perkins' next Enterprise AI hit — Joubin Mirzadegan, Roadrunner
Disclosure
Shulman: Suno borrowed significant code from Andrej Karpathy's nanoGPT for Bark
“There's a big shout out to Andre Karpathy's nano GPT. You know, there's a lot of code borrowed from there.”
Mikey Shulman Mar 14, 2024 ▶ 16:46 Making Transformers Sing - with Mikey Shulman of Suno

Every mention by year

tap a year for its mentions
00251350252023202420252026episodesmentions
013252023202420252026episodes it came up in
001.3132.5252023202420252026episodesmentions per episode
2026 27 mentions in 12 episodes 2 per episode
2025 49 mentions in 23 episodes 2 per episode
2024 27 mentions in 11 episodes 2 per episode
2023 7 mentions in 5 episodes 1 per episode

Played on the show (1)

Episodes where a recording of Andrej Karpathy was played rather than Andrej taking part, or where the tape carries an address with nobody putting questions to them. Listed because the words are on the record, kept out of every score on this page because they were not said on this show. We read this off the tape: who was spoken to, who was asked something, who answered whom.

EpisodeDateOn tapeWhat it is
llm.c's Origin and the Future of LLM Compilers - Andrej Karpathy at CUDA MODE Sep 21, 2024 18m aired address
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 200 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more