Adam Brown – Einstein's happiest thought: General Relativity from scratch
Adam Brown explains Einstein's General Relativity from first principles, showing how the equivalence of inertial and gravitational mass led Einstein to conceptualize gravity as curved spacetime rather than a force. The lecture progresses from special relativity through the geometric interpretation of gravity to black holes, demonstrating GR's explanatory power across vastly different scales.
Summary
Adam Brown, a physicist at Google DeepMind, delivers a lecture on General Relativity aimed at making this complex theory accessible to non-specialists. He begins by establishing that Einstein's special relativity—built on the principle that nothing travels faster than light—was incomplete because it didn't adequately describe gravity. The central problem was that Newton's inverse-square law of gravity appeared to violate special relativity by implying instantaneous gravitational influence.
Brown identifies Einstein's key insight: the equivalence principle. Einstein noticed that gravitational mass (how much an object is pulled by gravity) equals inertial mass (how much an object resists acceleration). This equivalence is merely coincidental in Newtonian physics but profound in suggesting gravity might be an inertial force—a force experienced when one is not moving in a straight line. This led to Einstein's revolutionary idea that gravity is not a force but rather a consequence of curved spacetime.
Using the analogy of airplane maps and great circles on Earth, Brown explains that what appears curved on a flat representation is actually straight on the curved surface itself. Similarly, objects in free fall (like chalk tossed in air or astronauts in orbit) follow straight lines through curved spacetime, while stationary observers experience fictitious gravitational forces.
Einstein's field equations mathematically describe how matter curves spacetime and how that curvature determines how matter moves—expressed in the slogan 'matter tells spacetime how to curve; curved spacetime tells matter how to move.' This theory successfully predicts mercury's orbit, light bending around the sun, and describes phenomena from falling apples to the expansion of the universe.
Brown then explores black holes—objects so dense that their escape velocity exceeds the speed of light. He derives the Schwarzschild radius (2GM/c²) using Newtonian reasoning, then shows how General Relativity refines this. Key concepts include gravitational time dilation (time passes slower in stronger gravitational fields, confirmed by atomic clock experiments and GPS), gravitational redshift (light loses energy climbing out of gravity wells), and the event horizon (a point of no return where escape becomes impossible).
A remarkable application is that lowering matter toward a black hole allows extraction of up to 100% of its rest mass energy—compared to ~10^-10 for chemical reactions or ~10^-2 for nuclear fusion. This is because gravity, uniquely, couples to all forms of energy. Brown explains that from an external observer's perspective, falling objects appear to slow and redshift as they approach the event horizon, never quite crossing it; but from the falling observer's perspective, crossing is unremarkable (for sufficiently large black holes), though ultimately fatal when reaching the singularity.
Historically, black holes were not believed to exist until theoretical work by Penrose and Hawking showed their formation is generic, and experimental evidence accumulated through orbital observations (stars orbiting Sagittarius A*), gravitational wave detection (LIGO's 2015 observation of merging black holes), and the Event Horizon Telescope's radio imaging.
Brown concludes by reflecting on General Relativity's remarkable span—from thought experiments about elevators to describing the entire universe's expansion. He notes that modern AI systems might discover physics similarly by exploring consistent mathematical frameworks with minimal empirical constraint, though he remains optimistic that AI will produce human-comprehensible explanations rather than inscrutable proofs.
About this episode
<p>Adam Brown is back!</p><p>General relativity is said to be the most beautiful idea the human mind has ever produced. Most of us will never get to fully appreciate its elegance by taking the 20-lecture graduate course Adam taught on it at Stanford. But in this episode, Adam distills the key idea at its heart so clearly and compellingly that even I could keep up lol.</p><p>At the core of general relativity, Einstein is trying to figure out the principle behind a particular coincidence: that the mass that resists acceleration and the mass that gravity pulls on just happen to be exactly the same. Adam then leads us through the path of insight which Einstein called his “happiest thought.”</p><p>Then Adam lectures on black holes. First, by showing how even under special relativity you could create a perpetual motion machine if black holes weren’t truly black. And then, by explaining why the observations of an infalling observer and a distant bystander to the black hole would be so radically different</p><p>Adam leads Blueshift, the team at Google DeepMind cracking science and reasoning, which gave us the opportunity to discuss at the very end how close we are to AIs that could rediscover general relativity from scratch. Stay till the close for some philosophy of science.</p><p>Watch on <a href="https://youtu.be/QbdbAhaJoCQ" target="_blank">YouTube</a>; read the <a href="https://www.dwarkesh.com/p/adam-brown-gr" target="_blank">transcript</a>.</p><p><strong>Sponsors</strong></p><p>* <a href="https://janestreet.com/dwarkesh" target="_blank">Jane Street</a> has traders from all sorts of different backgrounds. For example, I recently got to speak with Jed Thompson, a trader who started his career in particle physics. Jed told me how the habits he built as a physicist (like never running a calculation without first having a good guess at the answer) helped him build good trading intuition. So no matter what field you’re working in right now, your experience may be more applicable than you think. Check out open positions at<a href="https://janestreet.com/dwarkesh" target="_blank"> janestreet.com/dwarkesh</a></p><p>* <a href="https://crusoe.ai/dwarkesh" target="_blank">Crusoe</a> gave me early access to their new serverless fine-tuning product, so I decided to try fine-tuning a Dwarkesh-style question generator. Crusoe made this really easy: I just turned my interview transcripts into training data and then kicked off a run – I never had to touch infra or tweak hyperparameters. After training was done, I ran a blind eval with my team: they preferred the fine-tuned model’s proposed questions over my own suggestions about 30% of the time. Serverless fine-tuning goes live next week. Learn more at<a href="https://crusoe.ai/dwarkesh" target="_blank"> crusoe.ai/dwarkesh</a></p><p>* <a href="https://cursor.com/dwarkesh" target="_blank">Cursor</a>’s iOS app lets me kick off real work no matter where I am. For example, recently I was at dinner with friends when I had an idea about how to investigate the past few years of progress in sample efficiency. I pulled out the Cursor app, dumped my thoughts into a voice note, and 15 minutes later, Cursor had cloned the relevant repo, done the necessary analysis, and written up its findings. And now I’m expanding that work into a full write-up. Without the Cursor app, the idea would’ve floated away. Check out the app now at<a href="https://cursor.com/dwarkesh" target="_blank"> cursor.com/dwarkesh</a></p><p>Timestamps</p><p>(00:00:00) – The coincidence that led Einstein to general relativity</p><p>(00:16:42) – Gravity is a consequence of curved spacetime, not a force</p><p>(00:31:46) – Why black holes prevent unlimited energy extraction</p><p>(00:47:12) – Black holes are the ultimate power plants</p><p>(01:13:50) – The three ways we know black holes are real</p><p>(01:18:51) – How do we know black holes exist but not wormholes?</p><p>(01:24:21) – The first time we saw gravity bend light</p><p>(01:29:33) – How far can AI get without experimental evidence?</p> <br /><br />Get full access to Dwarkesh Podcast at <a href="https://www.dwarkesh.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4">www.dwarkesh.com/subscribe</a>
Key Insights
- Einstein recognized that Newton's law of gravity is mathematically inconsistent with special relativity because it implies gravitational influence propagates instantaneously, faster than light
- The equivalence principle—that gravitational mass equals inertial mass—was merely an observed coincidence in Newtonian physics but became a profound clue that gravity might be fundamentally different from other forces
- Unlike electromagnetism (mediated by spin-1 photons), gravity is mediated by spin-2 particles (gravitons), which is why gravity is attractive while like-charges repel and why gravity cannot be treated exactly like electromagnetism
- Einstein's central insight was that gravity is an inertial force, meaning objects in free fall are moving along straight lines through curved spacetime, while stationary observers (who experience gravitational force) are actually moving along curved paths
- The mathematical description of how spacetime curves is given by Einstein's field equations, which state that the presence of mass and energy causes spacetime curvature, and conversely, spacetime curvature determines how matter moves
- General Relativity successfully reduces to Newtonian physics at large distances and low gravitational fields, explaining why Newton's theory works so well for everyday phenomena
- A black hole forms when matter becomes so compact that its Schwarzschild radius (2GM/c²) exceeds its physical size, creating an event horizon from which not even light can escape
- From an external observer's perspective, objects falling into a black hole appear to slow down and become redshifted as they approach the event horizon, and never visibly cross it; but from the falling observer's perspective, crossing the event horizon is locally unremarkable
- Gravitational time dilation—the slowing of time in stronger gravitational fields—is now experimentally confirmed to extraordinary precision in GPS systems and atomic clock experiments
- Lowering matter toward a black hole's event horizon allows extraction of essentially 100% of its rest mass energy (mc²), making black holes theoretically perfect power plants, compared to chemical reactions extracting ~10^-10 of rest mass energy
- General Relativity was not widely accepted until Sir Arthur Eddington's 1919 solar eclipse expedition confirmed that light bends by twice the Newtonian prediction, making Einstein a global celebrity
- Modern large language models might discover new physics through exploring consistent mathematical frameworks with minimal empirical input, similar to how Einstein derived General Relativity, but Brown expects AI to produce human-comprehensible explanations rather than inscrutable proofs
Topics
Transcript
I'm back with Adam Brown. You currently need BlueShift at Google DeepMind, which is cracking science and reasoning. In a previous life, Adam was a prolific physicist, taught at Stanford, and did research on everything from cosmology, to string theory, to general relativity. It's said that general relativity is the most beautiful thing that human mind has ever conceived or seen. I was curious if there's a way that ordinary people like me could understand what is happening or have some vintage on why it's beautiful without taking your 20-lecture graduate course. So that was the prompt for this lecture, and I appreciate you being willing to do it. Super exciting to be here. And yes, I think the answer…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Dwarkesh Podcast
Grant Sanderson – AI and the future of math
The discussion centers on the rapid advancements of AI in mathematics, exploring its implications for the future of math and related fields. The conversation highlights how AI's capabilities impact traditional mathematical roles, the process of knowledge creation, and the potential for new insights in various domains.
The next big breakthrough will be AIs learning on the job
The speaker discusses how AI labs are betting on reinforcement learning from verified rollouts (RLVR) to achieve AGI, but argues this approach has fundamental limitations. He contends that true general intelligence requires continual on-the-job learning through weight updates, which current scaling paradigms don't adequately address.
The data black hole at the center of AI
The transcript argues that AI's primary driver of progress is data quantity and quality rather than architectural improvements or scaling, highlighting a massive gap in sample efficiency between humans and AI models. The speaker contends that current AI systems are fundamentally different from human intelligence, requiring orders of magnitude more data to learn skills. Despite this inefficiency, AI can still automate white-collar work due to the economics of scale and parallelism.
Ada Palmer – Machiavelli is the most misunderstood thinker of all time
Ada Palmer discusses Machiavelli's political theories and their historical context, emphasizing the instability of Italian city-states and the influence of the papacy. She explores how Machiavelli's personal experiences and insights shaped his writings, particularly in 'The Prince' and 'Discourses on Livy'.
Alex Imas and Phil Trammell – What remains scarce after AGI?
Economists Alex Imas and Phil Trammell discuss what will remain scarce after AGI, covering labor share stability, the 'relational sector,' wealth redistribution mechanisms, and implications for developing countries. They explore historical parallels to industrial automation, the plausibility of various economic scenarios, and why negative economic growth from AI abundance is theoretically very difficult to achieve.