Some talks change how you see a field. Whether you have twenty minutes or a free evening, these lectures and talks offer clear explanations of how AI works, where it came from and where it may be going. We list categories of talks with notable examples you can find easily on YouTube or official sites.

How language models work

Andrej Karpathy — “Intro to Large Language Models” (2023). An hour-long, non-technical explanation of what LLMs are, how they are trained (pre-training and fine-tuning), how they use tools, and their security challenges such as jailbreaks and prompt injection. One of the best single introductions to the topic.

Andrej Karpathy — “Deep Dive into LLMs like ChatGPT” (2025). A longer, comprehensive walkthrough of the full training pipeline, from internet data to reinforcement learning, aimed at a general audience.

3Blue1Brown — Transformers and attention chapters. Visual explanations of how attention lets models use context.

The history and foundations

Nobel lectures (2024). The Nobel Prize lectures of Geoffrey Hinton and John Hopfield (physics) and of Demis Hassabis, John Jumper and David Baker (chemistry) explain foundational work on neural networks and protein structure prediction in the laureates’ own words.

Turing Award lectures. Hinton, LeCun and Bengio gave lectures after receiving the 2018 Turing Award, outlining the history and future of deep learning from their perspectives.

MIT 6.S191 introductory lecture. A compact overview of deep learning fundamentals, updated each year.

AI in science

Talks by the AlphaFold team and others explain how AI predicted protein structures and how it is used in weather forecasting, materials discovery and mathematics. They show AI as a scientific instrument rather than just a chatbot.

Safety, ethics and society

  • Talks by researchers on algorithmic bias, such as Joy Buolamwini’s widely viewed TED talk on bias in facial recognition.
  • Lectures and interviews on AI alignment and risk by researchers including Geoffrey Hinton, Yoshua Bengio and Stuart Russell, who has also given public lectures — notably the BBC Reith Lectures in 2021 — on living with artificial intelligence.
  • Talks presenting skeptical or contrasting views, such as Yann LeCun’s lectures on the limits of current language models and the need for world models.

Watching contrasting views side by side is one of the best ways to form your own opinion.

Economics and the future of work

Lectures by economists on AI’s effects on productivity, labor markets and inequality offer grounded perspectives beyond technology hype. University economics departments and policy institutes publish many of these online.

How to get more from a talk

  1. Write down three key ideas after watching.
  2. Look up one term or paper mentioned that you didn’t know.
  3. Discuss it with a colleague or community.
  4. Check the date: in AI, a talk from two years ago may describe capabilities that have since changed.
  5. Look for follow-ups from the same speaker.

Build your own playlist

Combine one “how it works” talk, one history talk, one safety talk and one economics talk. Together they give a balanced foundation for understanding AI news and debates.

Talks for specific audiences

  • Managers and executives: talks on AI strategy, adoption case studies and economic impact.
  • Educators: talks on AI in classrooms, assessment and AI literacy.
  • Creators: talks on generative tools, rights and creative workflows.
  • Developers: conference talks on building reliable AI applications, evaluation and agents.

Watching critically

Ask what evidence each speaker offers, what incentives they may have, and which claims have been confirmed or overtaken by later developments.

What to watch next

  • New lectures explaining reasoning models and agents.
  • Public lecture series on AI from universities and broadcasters.
  • Recorded talks from major AI conferences.
  • Updated introductions as model capabilities change.

Key takeaways

  • Karpathy’s LLM talks are among the best introductions to language models.
  • Nobel and Turing Award lectures explain AI’s foundations from its pioneers.
  • Talks on bias, safety and risk show the societal stakes.
  • Watching contrasting views helps you form your own opinion.
  • Check talk dates, since AI capabilities change quickly.

Frequently asked questions

Where can I find these talks?

Most are on YouTube or the official websites of universities, the Nobel Prize and conferences.

Do I need a technical background?

Many of the listed talks, such as Karpathy’s introduction and Nobel lectures, are designed for general audiences.

Which talk should I watch first?

Karpathy’s “Intro to Large Language Models” is a good first choice for understanding today’s AI.