Susskind, Leonard

professor of theoretical physics at Stanford University; founding director of the Stanford Institute for Theoretical Physics; a father of string theory

Leonard Susskind: Quantum Mechanics, String Theory and Black Holes

Lex Fridman Podcast · Sep 26, 2019

• Leonard Susskind
• string theory
• quantum gravity
• quantum computers
• entanglement and observation
• emergent time and space
• black holes
• eternal inflation
• machine learning and consciousness
• arrow of time

Core ideas

Susskind credits his friendship with Richard Feynman not with changing his fundamental approach to physics but with validating it: both favored an intuitive, visualization-first method (closing one's eyes and picturing the phenomenon before reaching for equations) over purely symbolic, mathematically-driven reasoning, and seeing Feynman succeed this way gave Susskind confidence that his own less formally mathematical style could work. He reflects that modern physics — quantum mechanics, general relativity, quantum field theory — is inherently unintuitive relative to the classical intuitions humans evolved for (throwing stones, watching water), but that sustained exposure genuinely "rewires" physicists' intuition to the point where some of his colleagues find quantum reasoning easier than classical reasoning; he doubts, though, that humans can ever become fully, naturally comfortable visualizing four or more spatial dimensions, since our neural wiring seems fundamentally built for three, and higher dimensions are always handled through learned mathematical tricks rather than direct visualization.

Susskind explains why genuine quantum computers are categorically different from classical computers simulating quantum equations: a classical computer can solve the Schrödinger equation but is not itself functioning as a quantum system, whereas a quantum computer is an actual quantum-mechanical system directly instantiating the same physical rules (including the uncertainty principle) as whatever it's simulating — illustrated by the fact that storing the full quantum state of just 400 entangled qubits classically would require more information than could be physically stored in the entire universe, yet a 400-qubit quantum computer trivially represents that same state by simply being it. He's skeptical that quantum computers will provide broad exponential speedups across most computational problems (citing factoring as likely a rare, almost "fluke" special case), and argues their greatest genuine value will be simulating other quantum systems — chemistry, solid-state physics, materials science, quantum gravity — where classical computation becomes intractable and researchers instead build a working quantum analog they can slow down, probe, and manipulate in ways the original system doesn't allow.

He states that, based on conversations with the (few) neuroscientists he knows, the consensus view is that the brain functions classically rather than relying on quantum phenomena like entanglement or coherent superposition — a conclusion he says he'd romantically prefer to be false but suspects is likely correct — while noting that macroscopic quantum effects genuinely do exist in engineered materials like superconductors and topological insulators, where large-scale material properties depend fundamentally on quantum mechanics. On consciousness and introspection, Susskind argues human self-reflection is likely to give systematically wrong answers about how the mind actually works, citing neuroscience findings (e.g., dedicated neural circuits for detecting vertical versus 45-degree lines) that feel nothing like how thinking subjectively seems to operate; he speculates that machine learning — particularly systems that evolve not just their learned parameters but their own architecture, partly outside direct human control — may eventually reveal genuine mechanistic clues about consciousness that pure introspection cannot, citing self-play chess and Go engines (which developed strong play purely by playing against themselves, with no human teaching the strategy) as evidence that intelligence-like capabilities can emerge through evolutionary-style processes without being explicitly engineered.

On string theory, Susskind reframes it not as a self-contained subject with dedicated "string theorists" but as a mathematical tool within the broader project of theoretical physics — originally developed to explain the vibrational and rotational behavior of hadrons (protons, neutrons, mesons) in the late 1960s/early '70s, a use that was experimentally validated and is now settled; the same mathematics later resurfaced as a candidate theory of gravity operating at scales roughly nineteen orders of magnitude smaller than a proton, far beyond direct experimental reach. He argues string theory's single biggest contribution to physics has been demonstrating, with mathematical rigor, that quantum mechanics and gravity are not fundamentally inconsistent with each other — resolving a serious open worry (notably raised by Stephen Hawking) that once dominated the field — even though the theory cannot be tested in a laboratory at its relevant scale. Asked whether something deeper underlies quantum mechanics itself, Susskind discusses physicist Gerard 't Hooft's view that a deterministic, classical substructure might underlie and give rise to quantum mechanics (including the wave function and uncertainty), and states his own position is one of explicit epistemic humility: he sees no evidence for anything beneath quantum mechanics, but declines to claim it must be the final, bottom-level theory of reality.

Susskind connects entanglement directly to the physical meaning of "observation": an observer, technically, is simply a system with enough degrees of freedom to record information and become physically entangled with whatever it measures — a mathematically well-defined process he finds satisfying, even while admitting he remains personally uneasy with it in the same way he's uneasy visualizing five dimensions. He distinguishes space, which physicists increasingly believe is emergent (arising from entanglement and other underlying structure), from time, which remains built directly into fundamental physical equations as a symmetric, reversible parameter — with a genuine "arrow" of time only emerging at the thermodynamic level, once systems become large enough for entropy and statistical mechanics to be meaningful concepts; he notes that time-reversal is not merely theoretical but has been physically demonstrated for small systems (a handful of billiard balls can, with extraordinary precision, be made to reverse into their original configuration), while doing so for larger systems (thousands or millions of particles) becomes practically impossible due to compounding chaos, not any fundamental physical law — a distinction he's careful to separate from science-fiction "time travel."

On the origin of the universe, Susskind states his favored theory is eternal inflation, which he believes — properly understood — implies time extends infinitely in both directions rather than having a discrete beginning, and expresses skepticism toward the intuitive human preference for a definite starting point. Asked about the first black hole image from the Event Horizon Telescope and LIGO's gravitational wave detections, he calls them a magnificent scientific triumph confirming Einstein's general relativity works precisely as predicted even in black hole collisions, though he doesn't expect the imaging itself to reveal fundamentally new physics about black holes so much as new information about cosmic structure formation. He closes by naming what he considers a genuinely real but likely permanently unanswerable question: whether some intelligent agent underlies or is "responsible" for the universe as a whole, and whether such an agent would itself be subject to the laws of physics and quantum mechanics — a question he considers meaningful despite acknowledging that some philosophers would deny it qualifies as a real question at all, since it may not be answerable by any known method.

Excerpts