Quotesby GlyphAndrej Karpathy quote, Dec 19, 2025
Founder, Eureka Labs · Dec 19, 2025 10 months ago · Blog, 2025 LLM Year in Review
We're not "evolving/growing animals", we are "summoning ghosts".
Words found on the source page on Nov 6, 2026.
More from Andrej Karpathy
- And I've vibe coded entire ephemeral apps just to find a single bug because why not - code is suddenly free, ephemeral, malleable, discardable after single use. Vibe coding will terraform software and alter job descriptions.
- Anthropic got this order of precedence correct and packaged CC into a delightful, minimal CLI form factor that changed what AI looks like - it's not just a website you go to like Google, it's a little spirit/ghost that "lives" on your computer.
- LLMs are emerging as a new kind of intelligence, simultaneously a lot smarter than I expected and a lot dumber than I expected.
- Training on the test set is a new art form. What does it look like to crush all the benchmarks but still not get AGI?
- Software 1.0 easily automates what you can specify. Software 2.0 easily automates what you can verify.
- The more a task/job is verifiable, the more amenable it is to automation in the new programming paradigm.
- I do, but it’s business as usual because we’re in an intelligence explosion already and have been for decades.
- If you talk to a billion parameter model, I think in 20 years, you can have a very productive conversation.
- In my mind, education is the very difficult technical process of building ramps to knowledge.
- In my mind, this is more accurately described as the decade of agents.
- Reinforcement learning is a lot worse than I think the average person thinks.
- The reason you don’t do it today is because they just don’t work.
- The way I like to put it is you’re sucking supervision through a straw.
- We’re building ghosts or spirits or whatever people want to call it, because we’re not doing training by evolution. We’re doing training by imitation of humans and the data that they’ve put on the Internet.
- What takes the long amount of time and the way to think about it is that it’s a march of nines.
- As far as I can tell, the average person has never experienced a technological unlock this dramatic, this fast.
- Mary, Jim and Joes are experiencing the majority of the benefit, not Google or the government of the United States.
- So it strikes me as quite unique and remarkable that LLMs display a dramatic reversal of this pattern - they generate disproportionate benefit for regular people, while their impact is a lot more muted and lagging in corporations and governments.
On the same topics
- An AI with more real layers can actually be “smarter” in the conventional sense of the term.
- For 100 steps in a row, the AI can shoot delicate subtle ideas from layer to layer at the speed of light. Then there’s one step where it has to encode them into twenty-six glyphs invented by Phoenician turquoise miners in 1800 BC. Then it has to re-encode the Phoenician glyphs into delicate subtle lightspeed ideas before it can do anything else.
- The science of reading these thoughts is a subfield of AI interpretability, which is still in its infancy.
- This trains the AIs to be focused on task success, which naturally risks including things like reward-hacking, cheating, and single-minded pursuit of stated goals at the expense of ethical injunctions.
- It is important to not confuse massive steps in inference-time scaling, a dynamic which should be fairly predictable, with being the outputs of RSI, which is highly uncertain.
- If you have five hours, you’re probably going to do a lot better. The AI models are pretty similar.
About this quote
- Said by
- Andrej Karpathy, Founder, Eureka Labs
- Where
- 2025 LLM Year in Review
- Topics
- Research
- Company
Eureka Labs
- Source
- karpathy.bearblog.dev/year-in-review-2025/
A quote is one moment, not the whole argument. Read it in context at the source.