The Beginning of Infinity

In some books I look for a philosophy of life: guidance on living more truthfully, happily and usefully. The Beginning of Infinity by David Deutsch gave me a piece of that puzzle. Deutsch challenges familiar pictures of the world, including “Spaceship Earth”, and keeps asking for better explanations.

What struck me most was his case for optimism. Deutsch and I believe the world can get much better, even infinitely better. Problems created by today’s solutions call for new knowledge and further solutions. That has changed how I think about climate change: alongside reducing our impact, we can work on new solutions.

The chapter notes below come from my original reading, including the questions I was still thinking through at the time. The discussion of AI in particular reflects that earlier reading.

Opening notes: explanations

Deutsch describes scientific theories as explanations of what is out there and how it behaves. They are conjectures, or guesses, rather than conclusions derived only from observation. That helps explain how science can make predictions about things we have never experienced.

Good science comes from a tradition of criticism rather than reliance on authority. A theory must be testable: it should make predictions that an experiment could contradict. When a theory conflicts with reality, solving the problem means creating a better explanation. A good theory makes it harder to fool yourself.

One useful idea is “reach”: an explanation can solve problems beyond those it was created to address. That gives knowledge a value we cannot always anticipate when we first develop it.

Chapter 2: Reality and interpretation

Almost everything we experience comes at the end of a chain of interpretations. What we see in our minds is already an interpretation of what is out in the world. We get closer to reality by finding and correcting misconceptions in our theories, rather than treating our first impressions as reliable access to it.

Chapter 3: People, resources and problems

Many older beliefs put humans at the centre of everything. The principle of mediocrity pushes in the other direction, treating our place in the universe as unremarkable. Deutsch thinks this goes too far. Humans have unusual capabilities, and the place we live is unusual too.

Earth was not made for us and is not automatically hospitable. Try spending a European winter outside without clothes. Yet we live near a star, surrounded by useful matter, rather than in the emptiness of deep space. What makes the difference is our ability to find solutions. Other species also face disaster and death, and most species that have existed are extinct. If we are on Spaceship Earth, we are also its designers and builders.

Deutsch’s optimistic claim is that we can achieve anything with the resources at hand unless the laws of nature forbid it. Living on another planet, for example, depends on the habitat we can build, rather than whether its atmosphere happens to suit us already. Even our need for that particular atmosphere might eventually change. These possibilities are far away, but knowledge, once discovered, can keep being used. Deutsch also raises the possibility of extending our lives, a topic explored in Ending Aging.

The ingredients are matter, energy and evidence with which to test our explanations. Two of Deutsch’s maxims stayed with me: problems are inevitable, and problems are soluble. They changed how I thought about climate change. Alongside preventing damage and limiting our impact, we can develop solutions that make the planet more hospitable for humans.

Chapter 4: Evolution and knowledge

Human knowledge grows through variation and selection: we make conjectures, then criticise and test them. Deutsch connects this with evolution. Good adaptations, like good explanations, are hard to vary while still fulfilling their functions.

I found his argument against creationism convincing. He makes the gene-centred case for evolution: what spreads through a population matters, rather than assuming evolution works for the benefit of a species or an individual. There is more on that in The Selfish Gene. Ideas and genes are both replicators, but human knowledge can also explain things and have reach beyond its original purpose.

Chapter 5: Emergence

Higher-level abstractions can be partly autonomous from the smaller parts that make them possible. Deutsch calls this emergence. Understanding those abstractions requires more than reducing everything to its underlying components. Knowledge creation itself depends on emergent phenomena, even though it physically consists of processes at lower levels.

I found pages 118–121 somewhat convoluted and marked them to read again.

Chapter 6: Universality

A small change in a system can make it universal. Deutsch’s examples include language and number systems. Tally marks only take us so far, whereas our positional number system lets us represent numbers without inventing a new symbol for each one. Movable-type printing is another example; I wondered whether 3D printing might offer a similar jump.

Computers are universal machines for many purposes. My notes ask whether artificial general intelligence might one day emerge from such a system. Error correction, in computers and humans, is part of what allows knowledge to keep growing. RNA and DNA offer an earlier example of universality.

What I took from the chapter was that universality can arrive through a small change. It is a jump in what a system can do, rather than simply a gradual improvement at its existing task.

Chapter 7: Artificial intelligence

Deutsch argues that artificial intelligence must be possible in principle because a universal computer can simulate any physical process. At the time of my original notes, I wondered why decades of work had not yet produced the jump to general intelligence, even though systems such as Watson and AlphaZero seemed to have considerable reach.

Two ideas stood out: programming something requires understanding it, and judging whether a system is genuinely intelligent depends on an explanation of how it works. That explanation might be at an emergent level rather than a description of every underlying operation.

As I understood Deutsch’s argument, an evolutionary algorithm can be clever without creating new explanatory knowledge as it runs. Its performance can still depend on the programmer’s creativity and a great deal of computation. I wanted to know which AI systems met the criteria he set out in this chapter. We did not yet understand how creativity worked, which left me with more questions to explore.

Chapter 8: Infinity and the laws of nature

The possibility of knowledge growing without limit depends on several kinds of universality. The laws of nature apply beyond the place and time in which we discover them. People can create explanations with similarly wide reach, and universal classical computers can carry out computations beyond the particular tasks for which they were built.

The chapter also explores mathematical infinity, including Hilbert’s infinite hotel. I found this fascinating; this video on the infinite hotel offers another way into the idea.

The apparent friendliness of the laws of nature to computation prompted another question in my notes: could we be in a simulation? That does not necessarily resolve anything. It could simply leave us with simulations all the way down.

Chapter 9: Optimism and prediction

I like Deutsch’s insistence on innovation and new explanations. Predictions of doom often assume we will keep doing the same things. History gives us reasons to question that assumption, because what we think and do changes what happens next.

He distinguishes prediction from prophecy. A prediction follows from a good explanation; a prophecy claims to know what is not yet knowable. When the possible outcomes are not yet known, neither are their probabilities.

Stopping innovation, including the development of new weapons, will not automatically keep us safe. We could already harm each other with fire and swords. Expanding our knowledge can make us better able to survive. We might be doomed unless we solve a particular problem in time, but that is also what people are working to do. Better-insulated houses and better weather models are everyday examples of knowledge helping us weather storms.

Deutsch’s principle of optimism is that all evils are caused by insufficient knowledge. Criticism and good explanations therefore matter enormously. He connects this with the Enlightenment, also discussed in Sophie’s World, and the possibility of sustaining that tradition rather than allowing it to end.

Chapter 10: Dialogue and explanation

This chapter uses a dialogue involving Socrates to explore what makes a good explanation. Its parts are hard to vary while leaving the explanation intact. As our explanations improve and converge on the truth, we also have a better chance of understanding one another.

Chapter 11: The multiverse

The multiverse is one of those ideas I find counterintuitive and hard to grasp. Television series such as The Man in the High Castle, Fringe and Counterpart make use of parallel worlds, but Deutsch is arguing for a particular interpretation of quantum physics.

Much of this chapter was beyond what I felt able to reproduce confidently. My understanding of his conclusion is that the physical world is a multiverse, with information flowing in partly autonomous streams called histories. What we call a universe is an emergent feature of that larger structure. Within a history, objects approximately obey classical physics, while quantum interference gives us evidence about the wider structure.

Fungibility was another unfamiliar idea: initially indistinguishable instances of objects can become different, producing what appears to us as randomness within a deterministic multiverse. Deutsch describes changes between discrete quantum values in terms of interference and fungibility. I would need to spend more time with the argument to explain it well myself.

Chapter 12: Physics and philosophy

Deutsch describes us as emergent, partly autonomous flows of information within the multiverse. He is also critical of approaches to quantum physics that he regards as bad philosophy, because they obstruct the growth of knowledge rather than merely getting an answer wrong.

His description of happiness as continually solving one’s problems interested me. Unhappiness, in that account, involves not knowing how to solve them. The chapter also returns to the importance of explanation in science: treating a process as a black box, as in behaviourism, leaves something important unexplained.

Chapter 13: Choices and political systems

Deutsch explains why perfect political representation is not possible and why we should put more effort into creating new options. Rational decision-making involves explaining evidence as part of understanding the world, rather than only weighing the evidence for the choices already on offer.

A political system should let us remove bad choices. Deutsch argues that a two-party system can do this better in principle than one with more parties. I was less convinced by how that looked in practice. Watching American politics may have biased me, and he could still be right about the principle, but this was one argument I wanted to question.

Chapter 14: Beauty and explanation

Deutsch sees a connection between good explanations and beautiful things: both are hard to vary while preserving what makes them work. I found this an interesting account of beauty and thought I agreed with it. Elegance in an explanation can be a kind of beauty, and a deep truth can itself be beautiful.

He extends the argument to music, landscapes and the symmetry of a flower. He also argues that beauty can be universal across species, with signalling offering a way to think about that claim.

Chapter 15: Culture and memes

A culture consists of ideas that cause people to behave alike in some way. These ideas, or memes, replicate and change as they pass between people. Culture carries long-lived memes through time while also changing them.

Genes and memes are both replicators, but they spread differently and can have very different effects. A meme has to be interpreted, represented in someone’s mind and expressed again in behaviour. Creativity emerges at a higher level through conjecture and criticism.

Deutsch contrasts static and dynamic societies. In a static society, change happens too slowly for its inhabitants to notice, and customs, taboos and anti-rational memes resist criticism. In a dynamic society, ideas survive by revealing something true and useful, even when they change how people live. My notes connect this with earlier, smaller enlightenments and the possibility that our own might support a more enduring dynamic society.

Chapter 16: Creativity

Deutsch gives creativity a special place among biological adaptations because it can produce scientific and mathematical knowledge, art and philosophy. Its effects dominate our experience of the world. My original notes also raised his claim that, as far as we know, creativity evolved only once, in humans, alongside the question of AI.

In static societies, creativity can be used to recreate existing memes rather than develop new ones. Even copying an idea involves creating an interpretation of it: the recipient makes a conjecture, criticises and tests it, then tentatively adopts it.

Deutsch treats creativity as a property of software: it depends on how our brains work rather than a magical ingredient. My notes connect that capacity with both genes and memes.

Chapter 17: Progress and climate change

Deutsch argues for optimism about the problems we face, including climate change. We can work to fix them rather than worry ourselves to death. I can get behind that, while also believing we should lower our impact, including through technology. Let My People Go Surfing explores that side of the question.

We have to live with a solution and then solve the new problems it creates. For Deutsch, progress can continue indefinitely. That brings me back to his distinction between predictions and prophecies: we should not pretend to know future discoveries in advance. What I like most is the invitation to get on with finding solutions, rather than putting our heads in the sand or trying to press pause.

Chapter 18: The beginning of infinity

We are still at the beginning of infinity, with much more to learn and discover. I hope to see a great deal of it, including in the far future. Seeking good explanations through creativity and criticism gives us a way to keep going.

Still an inspiring book

Looking back in 2026, I still find this an inspiring book. Its confidence in our ability to create knowledge and solve problems has stayed with me.