Summary of "Андрей Карпати - Мы вызываем призраков, а не создаём живых существ"

High-level summary

Andrey Karpathy (former Tesla Autopilot lead) gives a pragmatic, wide-ranging discussion of current AI (LLMs, agents, reinforcement learning), how progress has actually happened, what’s missing, and how education should adapt. His core metaphor:

Today’s systems are “ghosts” — disembodied imitators trained on internet data — not “animals,” which are organisms shaped by evolution and lifelong embodied learning.

He expects steady, decade-scale progress rather than an abrupt one-year takeover.

Key technological concepts and analyses

Pretraining vs. evolution

Agents, LLMs, and the “Decade of Agents”

Reinforcement learning (RL) critique

Context window, KV cache, and working-memory analogy

Transformers as cortical analog / architectural trends

Model scaling, distillation, and the “cognitive core”

Collapse, synthetic data, and entropy

Multi-agent systems, culture, and self-play

Code models, auto-completion, and agents

Autopilot analogy and the “march of nines”

Economic and societal impact / takeoff views

Product features, guides, tutorials, and educational projects

NanoChat

Micrograd

CS231N and LM101N

Eureka (education project)

Practical learning advice

Industry tools & companies referenced

Research and engineering gaps / suggested directions

Practical takeaways / recommendations

Main speakers and sources

Category ?

Technology


Share this summary


Is the summary off?

If you think the summary is inaccurate, you can reprocess it with the latest model.

Video