Children will deploy all sorts of tactics to avoid going to bed. 鈥檚 youngest had a canny one: to distract his mum, an astronomer at the University of Edinburgh, UK, he asked her meaty questions about science. One night a few years ago, he posed one of the oldest chestnuts of all: could the reality we live in be a simulation?
For most parents, this would have been the cue to switch off the bedroom light. For Heymans, who is also Astronomer Royal for Scotland, the question was harder to ignore. Computer simulations are an integral part of her everyday work. Astronomers can鈥檛 slip a galaxy into the laboratory or wait a few billion years to see how one evolves, so they routinely build universes in code and test them against what telescopes reveal.
Her son鈥檚 question percolated in her mind during a trip to a telescope in Chile, where, beneath the black skies of the Atacama desert, it began to take on a life of its own. Eventually, Heymans . New 精东传媒 caught up with her to hear her unique approach to this conundrum, whether it is remotely possible to solve 鈥� and, crucially, whether her son was convinced by her answer.
Advertisement
Jacklin Kwan: I have to say, 鈥淒o we live in a simulation?鈥� is an awfully precocious question for a child to ask.
Catherine Heymans: I think nearly all parents find their children do tend to ask a lot of surprisingly deep questions. That is certainly my experience 鈥� and I don鈥檛 like not knowing the answers! That鈥檚 part of being a scientist, isn鈥檛 it? It became a game: they would try to get out of going to sleep by asking me trickier and trickier questions, because I鈥檇 get really into it and suddenly their bedtime would be missed because I was so excited.
Anyway, they struck gold. I was running off on an observing trip to the Paranal Observatory in Chile, so I was going to be gone for a long time, and they didn鈥檛 like it when I went away. They obviously thought: what鈥檚 the trickiest question we can ask her, so she has to stay and answer it?
My youngest had just started playing computer games, and they are really realistic. He asked: 鈥淐ould we be living inside a computer game?鈥� I couldn鈥檛 answer it before I had to run to the airport. Then I got to Paranal, which is an amazing, life-changing place. You are right out in the middle of the desert, with nobody else there, in pitch-black darkness. I found that, when I was outside having a break, with telescopes on either side of me, this question kept going through my head.
All this was quite some time ago, back in 2018, and I didn鈥檛 immediately do anything except think about the question. Then covid-19 hit and I got really ill. I鈥檝e had long covid for the past four years. At one point, I was so ill that I couldn鈥檛 get out of bed or tolerate light or sound, even though my brain still worked perfectly. All I could think was: 鈥淚f the computational overlord is real, they need to reboot me. I need to be taken to IT. I need more RAM. The computer simulation has run out of memory.鈥�
Because I was at home and couldn鈥檛 keep up with my research at the same pace, and because this existential question had been going around in my head for so long, I decided the only way to find an answer was to write a book.
This isn鈥檛 a new question 鈥� even Plato was thinking about whether reality was real. What was your way in?
The first question is: is this even a scientific question to ask? Some scientists argue that there鈥檚 no point asking a question that is impossible to answer. You are never going to be able to prove that we鈥檙e living in a computer simulation unless it鈥檚 a bad one. People talk about glitches [that would give it away], but if we were in a simulation bad enough to contain obvious glitches, we would already know about it. If this is a computer simulation, then it is a really, really good one.
But what if we take an indirect route? What if we ask: 鈥淲hat would I need to create a reality so perfect that I couldn鈥檛 distinguish it from true reality?鈥� What would you have to do if you wanted to become the computational overlord, the master of all worlds? If I could prove that this level of technological advance would be impossible to deliver, then we could all be happy in the knowledge that we are real!
I started with the idea of simulating light because I鈥檓 an astronomer, realising that it鈥檚 this electromagnetic force that is behind all of our senses 鈥� sight, touch, smell, taste and even sound 鈥� in principle, they鈥檙e all built on the same mathematics. Gravity came next. Albert Einstein鈥檚 theory of general relativity describes gravity and orbital motion really well, but it doesn鈥檛 describe dark matter and dark energy (that is still on my to-do list). Then you have to ask what we are made from, which takes you into quantum physics.
That鈥檚 an awful lot of coding already.
And it doesn鈥檛 stop there. A futuristic overlord would need some theory of everything that meshes all of that together to establish the rules to code reality at the quantum level at some advanced quantum-computing centre. When I did ballpark calculations for the book, however, I realised that, to simulate absolutely everything in the universe all the time, you鈥檇 have to have an almost infinite resource of energy 鈥� I think we can more or less rule that out.
Frustratingly, though, I wasn鈥檛 able to disprove the hypothesis that I鈥檓 part of a much smaller virtual world. There is a possibility that there鈥檚 just a wee bubble around me that is being fully simulated. I wasn鈥檛 able to rule out the idea that everything else I experience beyond the bubble could be driven by an energy-efficient AI illusion of the real world.
It wasn鈥檛 just about the how, though. You got interested in the why, too.
Yes, I also thought about why someone would want to create a simulation of conscious beings like us. Elon Musk鈥檚 argument is that many civilisations will eventually acquire the technological capability to create hundreds of thousands of simulated worlds, with billions of realisations played like computer games, over and over again throughout cosmic space-time. This would make the likelihood that we are in the base reality that will eventually go on to create these simulations very small, and we are therefore almost certainly simulated.
I disagree with him! If we were in an entertaining computer game, we would be driving motorcycles off cliffs every 5 minutes in epic fails. It doesn鈥檛 feel like there is something controlling me or constantly intervening. The fact that I have a conscious experience of reality and a sense of free will makes it very unlikely that this is a computer game.
But could this be a computer simulation? I think if we are in a simulation, it would be as part of a scientific experiment, in the same way that I create numerical simulations of galaxies and dark matter, or my colleagues create simulations of DNA. We use computers to inform us about the science of our reality. We code something, see how it compares with the real world, and that is the scientific method.
As an astrophysicist, I cannot take a galaxy into the office in my back pocket to prod and poke it in the laboratory 鈥� the timescales are too long, the distances are too wide, the conditions are too extreme. So, we have developed a different way of doing science. We say: 鈥淭his is what we think is happening in the universe. I will write the code to create that reality in my computer and compare it with what is out there. Oh, it is completely different. My ideas are wrong. Start again.鈥� And I think the reason you would build a simulation of our reality is to understand consciousness.
Why would it be consciousness that the simulators, whoever or whatever they may be, are studying?
If you make the question, 鈥淎m I living in a simulation?鈥�, the one truth that I know is that I exist as someone capable of conscious experience and thought. As philosopher Ren茅 Descartes wrote in 1637: 鈥淚 think, therefore I am.鈥� If this is a computer simulation, it has to contain that key ingredient.
One of the reasons why the so-called hard problem of consciousness is hard is because you can only ever know that you yourself are conscious. You can never find out for certain whether somebody else is, and no experiment you can design will directly test their conscious experience. But if you think consciousness arises out of the fundamental physics that we know, out of our biological nature and how our brains develop as we age and grow, then, in principle, it could be coded 鈥� and then you can carry out a scientific experiment.
Imagine that you could run multiple simulations of the birth and development of the same human, but in different settings. In one setting, you could create an environment where there is no music and then suddenly play some in their teenage years. How do they react? Or an environment where no one is funny: do they nevertheless develop a sense of humour?
You could even have a setting where no one in this simulated person鈥檚 environment exhibits any form of pain. Does your virtual test subject still wince when they hurt themselves? If they do, they are likely having a conscious experience of pain. If they don鈥檛, they are likely not conscious, only ever mimicking the behaviour around them in the other settings where everyone cries when they fall.
There are, of course, serious moral and ethical issues here. I鈥檓 not sure anybody should perform those experiments. But I can imagine a far-future scientist using controlled simulated environments to explore consciousness in ways that aren鈥檛 possible in our own reality.
You began this journey partly because of an existential need to know whether you are simulated. Did you find any closure?
I really enjoyed talking to David Chalmers, the philosopher who came up with the hard problem of consciousness. I emailed him out of the blue and said, 鈥淚鈥檓 the Astronomer Royal for Scotland. I鈥檝e got some questions about consciousness,鈥� thinking he would never reply.
He wrote back, amusingly addressing me as 鈥測our celestial eminence鈥�, and we had the most amazing conversation. He was very generous with his time. The thing he said that has really helped me is: if you feel real, then you are real. It doesn鈥檛 matter whether you are made up of quarks or whether you are made up of qubits, and that has been beautiful.
Was your son convinced by the answers in your book? I assume he knows about it.
I told him about it! He鈥檚 very disappointed in me that there isn鈥檛 a definitive answer, but he at least now knows how science really works and is willing to embrace the uncertainty of it all, and he insisted on closing the book with a parting message to our computational overlords.
It鈥檚 a poem that says: 鈥淓ven though there鈥檚 a very good probability / that you don鈥檛 exist / we nevertheless hope / you鈥檙e pleased with this book / that you鈥檝e inadvertently created.鈥� And that was our fond farewell to the computational overlords.
