Father Of Quantum Computing: How To Think Like A Genius - David Deutsch - #1164

I want to get past civilizational optimism and into what it does to a person: AI doom, human potential, and how you’ve changed since your last book. Do I get a free pass because I’m only human?

Unless you are a golden retriever, yes. Learning is not knowledge transferred into a passive mind. You make guesses about what is meant, then expose them to error. Failure to communicate belongs with the communicator too.

We hate being wrong. With partial knowledge and enormous ignorance, how do we keep convictions without making certainty a personality trait?

Don’t seek certainty; seek better error-correction. Status, intentions, and expertise do not decide a claim. Even an enemy can be right. What matters is whether an explanation survives criticism better than alternatives.

Nobody can re-derive reality before ordering lunch. We use credentials and reputations. How does this stay usable?

You act through theories of the situation. I buy the sandwich because I have an account of the shop, supply chain, regulation, and incentives—not because I bow to reputation. A small fact can overturn a large reputation.

The internet has collapsed the private scratchpad into the public lecture. People spot a miss and dismiss the wider thinker. Your Fermi-problem solution was to put speculative answers in fiction, a few million light-years from critics.

We do not know whether we are alone, and it is easy to invent reasons aliens have not appeared. That is fantasy against fantasy; get over it. Real optimism is not cheerfulness: failure means the relevant knowledge has not yet been created, not that progress is impossible.

That is agentic, but not rosy. A terminal diagnosis may reflect missing knowledge without giving a patient time to become the world expert and cure themselves.

Exactly. You cannot promise which problem you will solve. But reason never reaches a final wall with no purchase. Pessimism dresses as sophistication because despair is easy to write; a convincing rescue must actually solve the problem.

AI doom and climate catastrophe can become arbitrary selections from innumerable disasters. Fear of nuclear power grew so loud that societies retreated from it, with grave environmental consequences. Language models are extraordinary, but more compute is not an account of explanatory creativity.

Education still treats minds as buckets, then mistakes exam performance for understanding. Voluntary access to remarkable people can be valuable; compulsory subordination is harmful. Learning should resemble mountaineering: hardship chosen for a problem you love, not cramming for a test.

So there is a difference between a good hard path and gritting your teeth through misalignment. Meanwhile, we still have rubbish to take out.

Get the unpleasant thing over with, then improve it: a better bin, easier route, fresh air, a neighbour. There is nearly always some improvement available. Find a problem you love enough to live with; solved problems bring delightful problem children. I’m doing that with a quantum-theory textbook, a difficult nonfiction book, fiction, and my day job. Fingers crossed, one may arrive within a year.