learning by doing

3Blue1Brown's visual walkthrough of the attention mechanism inside transformers, covering self-attention, multiple heads and cross-attention. It is chapter 6 of his deep learning series.

Engineering Explained compares Lexus's steer-by-wire yoke on the RZ 450e with Tesla's yoke. It covers how steer-by-wire works and how variable steering ratios address the yoke's flaws.

Musician and audio engineer Benn Jordan tries Dolby Atmos as a consumer, covering the theory of immersive audio, its business model, and home theater and headphone listening. He concludes it often falls short of the pitch.

3Blue1Brown's first chapter on deep learning explains what neurons and layers are and the math underlying a basic neural network, using handwritten digit recognition with visual animations.

3Blue1Brown's second neural network chapter explains cost functions and gradient descent, showing how a network learns by minimizing error during training.

Autoline's John McElroy and Caresoft president Terry Woychowski look at the BYD Seagull, an electric hatchback that sells in China for about $11,500. Woychowski points out the cost-saving design and manufacturing choices, and McElroy takes it for a short test drive.

LegalEagle analyzes Katko v. Briney, the Iowa case about a homeowner whose shotgun booby trap injured a burglar, and the limits on deadly force in defending property.

Veritasium debunks the idea of learning styles, such as being a visual learner, with an interview with Daniel Willingham and a memory experiment. It argues that a multimodal approach helps everyone.

3Blue1Brown's third deep learning chapter gives a visual, intuitive account of backpropagation and what happens to a neural network as it learns. A follow-up video covers the underlying calculus.

Rational Animations explains Robin Hanson's grabby aliens model, which argues that humanity looks suspiciously early in cosmic history. Expanding civilizations set a deadline that explains our timing and bears on the Fermi paradox.

Yosh trains a reinforcement learning agent to drive Trackmania until he can no longer beat it. The video explains how the training works and the design choices involved.

Illumina's 3D animation of its sequencing-by-synthesis workflow, from sample preparation and cluster generation to sequencing on a flow cell and data analysis.

Comedian Dominic Frisby gives a ten-minute explanation of land value tax. He argues it would fix problems of rent-seeking and the economics of land ownership.

3Blue1Brown's visual explanation of how transformers and GPT-style language models work, tracing data through tokens, embeddings and attention.

Matt Parker of Stand-up Maths demonstrates MENACE, Donald Michie's 1961 machine made of matchboxes that learns to play noughts and crosses through reinforcement. He builds and trains a physical version.

Ray Dalio's 30-minute animated video explains how the economy works through credit, debt, interest rates, and short- and long-term cycles. It is based on his template for understanding the economy.

An animated 44-minute summary of Ray Dalio's book Principles for Dealing with the Changing World Order. It lays out the Big Cycle of rising and declining nations, covering debt cycles, internal conflict, and shifts in global power over the past 500 years.

Arthur Keller gives a roughly half-hour crash course on environmental collapse, covering exponential growth, carrying capacity and ecological overshoot.

Veritasium examines which experts have real expertise, drawing on chess research and work by Klein and Kahneman. It explains that expertise comes from pattern recognition built through deliberate practice in valid, feedback-rich environments.

Techmoan's beginner introduction to the compact cassette, testing metal tapes and Dolby S noise reduction on a Sony TC-S1 deck. He checks whether a cassette can sound close to a CD when good tape and equipment are used.

Robert Miles explains reward modelling, in which an AI learns what humans want from their preference comparisons instead of a hand-written reward function. He walks through the 2017 paper on deep reinforcement learning from human preferences.

Two Minute Papers covers OpenAI's multi-agent hide-and-seek experiment, in which agents learned emergent tool use and found ways to break the game's rules. It summarizes the 'Emergent Tool Use from Multi-Agent Interaction' paper.

An animated adaptation of Captain David Marquet's talk on greatness, based on his book Turn the Ship Around!. It argues for pushing decision-making down to the people doing the work instead of controlling everything from the top.

3Blue1Brown's chapter 4 on deep learning gives a more formal, symbol-based account of backpropagation calculus, connecting the intuition from the previous chapter to the math found in textbooks and code.

A 1951 Bell System training film showing how telephone calls are routed through a step-by-step (Strowger) switching office.

Technology Connections takes apart the step-by-step electromechanical logic of a 1970s pinball machine. He shows how relays, steppers and switches track state and score without electronics.

A short TED-Ed animated lesson by Anita Collins on what happens in the brain when people play a musical instrument. It argues that playing engages many brain areas at once, like a full-body workout, with long-term benefits.

3kliksphilip summarizes Valve's GDC talk on VACNet, a deep-learning system that flags CS:GO cheaters and feeds the Overwatch review queue. It covers Trust Factor, VAC, and how the data is collected and analyzed.

A Veritasium explainer on how the brain works and how we learn. It is inspired by Kahneman's Thinking, Fast and Slow and covers System 1 and System 2 thinking and why harder learning builds skill.

A guest video by Welch Labs on 3Blue1Brown explaining how text-to-image and video models work. It covers CLIP, shared embedding spaces, diffusion models (DDPM, DDIM), conditioning, guidance and negative prompts.

Yuta asks whether native Japanese speakers can actually write kanji by hand, showing that many struggle to recall characters from memory.

Part 3 of Kirby Ferguson's Everything is a Remix series, using Apple and the iPhone to show how invention builds on copying and combining prior ideas. It argues that creativity and innovation are inseparable from copying.

A Climate Town video on how the auto industry reshaped American cities and culture through advertising, lobbying and campaigns like the 1939 World's Fair. It is snarky in tone but backed by a published source document.

A short Veritasium video explaining regression to the mean and how ignoring it leads to wrong conclusions about punishment versus reward, medical treatments and sports streaks. It uses the example of flight instructors who believed criticism works better than praise.

A short RepresentUs explainer on how money corrupts American politics and what an anti-corruption act could do about it. It covers lobbying, super PACs and campaign finance, and pitches the group's proposed reforms. The video was later superseded by an updated version.

Veritasium examines research suggesting science explainer videos like Khan Academy's make students feel they understand while tests show they have not learned, because misconceptions go unaddressed. It suggests how to partly fix this.

A short RepresentUs explainer arguing that US politics is shaped by money, citing the Gilens and Page Princeton study on how little average citizens influence policy.

Veritasium visits Chernobyl, Fukushima and other places to find who on Earth is exposed to the most ionizing radiation. The answer turns out to be surprising and puts radiation risk in context.

Matt Parker follows the community effort to speed up his slow Wordle-solving code, which ended up running hundreds of millions of times faster. He walks through the techniques contributors used.

Two Minute Papers covers a paper on speeding up fluid and smoke simulation with convolutional networks. It contrasts the approach with earlier regression forest methods.