Chris WilliamsonFather Of Quantum Computing: How To Think Like A Genius - David Deutsch
Chris Williamson
Fallibility over certainty 0:00
The discussion starts with a claim that when you are misunderstood, the fault usually lies with the communicator. That leads into Karl Popper’s idea that knowledge grows in the same broad way in science, mathematics, child language, and even evolution: people guess, test, and correct errors. You cannot download understanding. It comes from within the learner, so authority is a poor guide. We are all alike in our infinite ignorance, and treating oneself as an authority is a mistake.
Using theories in life 6:01
Strong convictions should rest on error correction, not on a feeling of certainty. Certainty can mislead, and so can trust in a person’s goodness, status, or reputation. A hostile person may still know the truth on one point, and an enemy warning may be worth believing. For everyday choices, you use theories instead of first principles. A sandwich shop, a doctor, or an airline is judged by background facts, regulation, and history, not by a name alone.
Criticism and timing 19:31
Ideas can be made up very easily, and they can be wrong just as easily. One way to test them is criticism, but it matters when you expose them. If you attack an early idea too soon, you may cut off a false line of thought that could still lead to something true.
Optimism and AI fears 24:01
Optimism is not cheerfulness. It means that failure usually comes from missing knowledge, and that going from bad to good is always possible in principle. That does not guarantee success in any one case. A terminal illness may still stay terminal, and reason may still not reach far enough. The same outlook is applied to fears about AI and climate catastrophe. These risks could happen, but they are not special barriers compared with all the other dangers people always face.
LLMs and creativity 37:31
LLMs can do striking things, but they do not have a module for creativity. They were built to speak English well, not to capture what creativity really is. By contrast, chess engines beat grandmasters by a very different kind of computation. You know how to write the engine, but not the grandmaster. That is why LLMs do not show AGI, and why more computer power alone is not enough.
Education and discipline 48:01
Education still follows the bucket theory, as if knowledge were poured into empty minds. That makes tests and exams look like the thing itself, when they are only poor proxies. The internet can weaken compulsory education and let you reach knowledgeable people directly. Discipline is not always grim self-denial. For a mountaineer, hardship is part of the chosen task, not something to suppress.
Discipline and desire 57:30
Mountaineering is used as the right model for learning. The climber accepts pain, boredom, fear, and despair because the goal matters and the work has been embraced. Formula 1 racing shows the same thing: a driver must react at about half the time of ordinary human reflexes, acting before full information arrives because the track and car are understood so well. That kind of discipline is praised when it serves a loved task. It is blamed when it means grinding through GCSE maths for a summer, scraping by, and forgetting everything the next day. A rational society would make more of life line up with what people want to do, so you rely less on sheer will and improve the small bad tasks around you.
AI-generated summary. It can be wrong or incomplete - check anything that matters against the original.
