Scott Aaronson: A professor at UT Austin and director of its Quantum Information Center, previously at MIT. He is a leading theoretical computer scientist in quantum computing and computational complexity, and a renowned science communicator.
What this episode covers
Scott Aaronson explains why technical scientists should care about big philosophical questions, framing them as solvable 'Q prime' sub-questions reachable through math and empirical work. He gives an accessible account of quantum computing built on amplitudes, superposition, and interference, clarifying that quantum computers do not simply try every answer in parallel. He walks through decoherence, quantum error correction, and why breaking cryptography remains far off due to the millions of physical qubits required. He breaks down Google's quantum supremacy result and its sampling-based verification, and warns sharply against hype, especially overstated claims about quantum machine learning speedups.
The guest's own work
The Ghost in the Quantum Turing Machine
Scott Aaronson
“III had this essay called the ghost in the quantum Turing machine it's you know one of the crazier things I've written”— guest
Why Philosophers Should Care About Computational Complexity
Scott Aaronson
“you described some examples of such Q prime sub questions in your long essay titled white philosophers should care about computational complexity”— Lex Fridman
Also referenced (named, not recommended)
The Idiot
Fyodor Dostoevsky
“I see myself in these conversations as the main character for one of my favorite novels badesti husky called the idiot”— Lex Fridman
Samsung Galaxy S20
Samsung
“I saw that Samsung announced the new Galaxy S 20 and of course right away Technium right home has a new episode that summarizes all that I needed to know”— Lex Fridman
Computing Machinery and Intelligence
Alan Turing
“what he did in his Computing Machinery and intelligence paper was exactly you know he explicitly started with the question can machines think”— guest
Big reveals from this episode
- Aaronson reveals Turing dropped his Cambridge course halfway through to go work at Bletchley Park.
- He explains a thousand qubits requires two-to-the-1000 amplitudes, more numbers than fit in the observable universe.
- He debunks the popular claim that quantum computers work by trying every possible answer in parallel.
- Breaking RSA cryptography would require millions of physical qubits, far beyond today's devices.
- Aaronson calls Andrew Yang's tweet that no code is uncrackable due to quantum computing 'premature'.
- His 18-year-old undergrad Ewin Tang proved a celebrated quantum machine-learning speedup was false by finding a classical algorithm.
- He warns much of the business case for quantum computing rests on 'extremely shaky foundations'.
Worth remembering
- The 1939 transcript of Turing arguing with Wittgenstein is one of the more fascinating documents Aaronson has read.
- Democritus described the hard problem of consciousness in 400 BC in terms recognizable today.
- Quantum mechanics is best understood as a generalization of probability allowing negative and complex amplitudes.
- A computer at 10-to-the-43 operations per second would collapse into a black hole, proving Moore's law must end.
- Each logical qubit needs roughly a thousand physical qubits with known error-correcting codes.
- Google's verification test required a roughly 9-quadrillion calculation on a classical supercomputer.
- The Haber-Bosch fertilizer process from a century ago is a many-body quantum problem no one fully understands.
- Grover's algorithm gives only a square-root speedup, not an exponential one, for AI and optimization problems.