Reality+: Virtual Worlds and the Problems of Philosophy

by David J. Chalmers

A cosmic simulation can simulate a universe in which our own galaxy never formed. An evolutionary simulation can simulate a version of Earth in which humans never evolved. A military simulation can simulate a world in which Hitler never invaded the Soviet Union. Eventually, a personal simulation might simulate what ...

thought expe...

an experiment you carry out simply by thinking. You describe a possible world (or at least part of one) and see what follows. Plato’s cave is a thought experiment. He imagines a world where prisoners can see only shadows cast on a cave wall, and asks how their lives compare to the lives of people outside the cave. Zhuangzi’s butterfly is a thought experiment. Zhuangzi describes a world in which he remembers d...

Thought experiments fuel science fiction. Like philosophy, science fiction explores the world as it could be. Any given science-fiction story is a thought experiment; the author conjures up a scenario and watches what follows. H. G. Wells’ The Time Machine conjures up a world containing a time machine and then lays out the consequences. Isaac Asimov’s stories in I, Robot conjure up a...

Ursula Le Guin’s classic 1969 novel The Left Hand of Darkness describes a possible world where humans on the planet Gethen have no fixed gender. As Le Guin puts it in her 1976 article “Is Gender Necessary?”: “I eliminated gender to find out what would be left.” In an introduction to the novel, she writes: If you like you can read [this book], and a lot of other science fiction, as a thought-experiment. Let’s say (says Mary Shelley) that a young doctor creates a human being in his laboratory; let’s say (says Philip K. Dick) that the Allies lost the Second World War; let’s say this or that is such and so, and see what happens. . . . In a story so conceived, the moral complexity proper to the modern novel need not be sacrificed, nor is there any built-in dead end; thought and intuition can move freely within bounds set only by the terms of the experiment, which may be very large indeed. Thought experiments yield many insights. Le Guin’s thought experiment gives us insight into a possibility: It tells us something about gender as...

Thought experiments can be far-fetched, but they often teach us something about reality. Le Guin says that in writing about gender she is “describing certain aspects of psychological reality in the novelist’s way, which is by inventing elaborately circumstantial lies.” Le Guin’s Gethenians may not exist, but aspects of their nature may resonate with the lived experience of many people, including some nonbinary people.

Asimov’s exploration of artificial intelligence in robots can advise us about how to interact with real AI systems once they’re developed.

Plato’s cave helps us to analyze the complex relation between app...

“I can’t help wondering,” mused Staghorn, “of whose computer we’re a part right now—slight factors in the chain of causation that started God knows when and will end . . .” “When someone pulls the switch,”

Nick Bostrom

“Are You Living in a Computer Simulation?,”

First, the simulation needs to be lifelong, or at least for as long as we can remember. Being in a simulation since yesterday doesn’t count. Second, the simulation needs to have been designed by a simulator. A computer program that popped up randomly without a simulator wouldn’t count. Both of these factors are part of the simulation hypothesis as people ordinarily think of it.

Sometimes people say that no universe can contain a perfect simulation of itself, since the universe would need a simulation of the simulation, and a simulation of that, and so on, leading to an infinite stack of simulations. Now, such a stack is not obviously impossible. Perhaps an infinite universe could devote a small fraction of its resources to running a (still infinite) simulation of itself. The resulting stack of simulations would be no problem for an infinite universe. Even a finite but expanding universe could run an ongoing simulation of the past that lags a little behind reality.

Were it indeed the case that no universe could simulate itself, that still would not rule out the simulation hypothesis. There’s no reason to suppose that the simulated universe and the simulating universe should be exactly the same.

we’re in a simulation, the simulating universe may have an entirely different physics from ours and ...

If the simulating universe is infinite, and has infinite resources, simulating a fini...

To sum things up, I would say that in principle we can get evidence for and against various imperfect simulation hypotheses, which will presumably have empirical consequences we can test. So these imperfect simulation hypotheses count as scientific hypotheses. They may not yet be serious scientific hypotheses, since we don’t yet have ...

However, we can never get experimental evidence for or against perfect simulation hypotheses. A nonsimulated world and a perfect simulation of it will seem exactly the same. So, according to the testability criterion, the hypothesis that we’re in a perfect simulation is not a scientific hypothesis. Instead...

Some hard-nosed scientists and philosophers may hold that because it’s untestable, the perfect simulat...

I’ll argue in chapter 4 that this is incorrect. In principle, we can construct perfect simulated worlds ourselves, with beings inside them. There will be no way for those beings ever to know that they’re in a simulation. The simulation hypothesis is demonstrably true of those beings. It follows that the hypothesis is meaningful. It may also be true of us, or it may not. Perhaps we ...

What about the original simulation hypothesis, saying that our world is a computer simulation? Is this a scientific hypothesis or a philosophical hypothesis? The philosopher of science Karl Popper insisted that the hallmark of a scientific hypothesis is that it is falsifiable—capable of being proved false using scientific evidence. We’ve seen that the simulation hypothesis is not falsifiable because any evidence against it could be simulated. So Popper would...

There can be scientific hypotheses—for example, about the early universe—that could never be falsified. But I’m happy to say that the simulation hypothesis is not a squarely scientific hypothesis but one that is partly scientific and partly philosophical. Some versions of it are subject to test, while other versions of it are impossible to test. But whether t...

What is the relationship between computer simulations and virtual worlds?

Recall that a virtual world is an interactive, computer-generated space.

Is every virtual world a simulation? Is every simulatio...

picture, the simulation hypothesis suggests we’re living in a fully immersive virtual world. A virtual world is immersive when you experience it all around you as if you were right there,

A virtual world is fully immersive when one is immersed in the virtual world with all of one’s senses, experiencing it just as we experience the physical world.

The ancient Greek philosopher Sextus Empiricus (second or third century CE) questioned our knowledge of science.

His Indian Buddhist contemporary Nāgārjuna questioned whether we gain knowledge from philosophy.

The 11th-century Persian philosopher al-Ghazali questioned our knowledge o...

The 18th-century Scottish philosopher David Hume questioned our kno...

The contemporary American philosophers Grace Helton and Eric Schwitzgebel have respectively questioned whether we know other people’s mi...

A common view of knowledge, going back to Plato, is that knowledge is justified, true belief. To know something, you have to think it’s true (that’s belief), you have to be right about it (that’s truth), and you have to have good reasons for believing it (that’s justification).

“Knowledge itself is power.”

Descartes

Meditations on First Philosophy.

His demolition crew included three classic arguments—concerning illusions, dreams, and demons—that cast doubt on our knowledge of the external world. These arguments weren’t entirely new. Illusions and dreams were standard fare for skeptics in ancient times such as Sextus Empiricus and the Roman orator Cicero, as well as for medieval thinkers such as the 5th-century North African saint Augustine and the Persian philosopher al-Ghazali.

We’ll see that demons were used by Descartes’s contemporaries as well. Nevertheless, Descartes gave these arguments their most influential formulation.

Descartes’s first argument was based on illusions. Our senses have deceived us before. How can we know...

No sensory illusion could give people the sense of having an entirely different body or being in an entirely different environment. He wrote: “Yet although the senses occasionally deceive us with respect to objects which are very small or in the distance, there are many other beliefs about which doubt is quite impossible, even though they are derived from the senses—for example, that I am here, sitting by the fire, wearing a winter dressing-gown, holding this piece of paper in my hands, and so on.”

body, and I’ll sense that it’s mine. Inside VR, Descartes could even have the sense that he’s sitting by the fire in a dressing gown, holding a piece of paper. VR thereby strengthens Descartes’s original argument based on illusions. Technology makes it harder to know that we’re not experiencing an illusion right now.

Descartes’s second argument was about dreaming: Dreams are like reality. How do we know we’re not dreaming?

Descartes thought the dream argument was stronger than the illusion argument.

Descartes’s third and most notorious argument is an argument about deception. An all-powerful being could deceive me completely, by giving me experiences of a world that does not exist. How do I know this isn’t happening to me?

Descartes talked of a genium malignum,

“bad genie”

While a benevolent God might refuse to deceive us, an evil demon would have no such compunctions.

I will suppose therefore that not God, who is supremely good and the source of all truth, but rather some evil demon of the utmost power and cunning has employed all his energies in order to deceive me. I shall think that the sky, the air, the earth, colours, shapes, sounds and all external things are merely the delusions of dreams which he has devised to ensnare my judgement.

16th-century Spanish theologian Teresa of Ávila wrote her own meditations in which deceiving demons played a central role. For Teresa, the issue was belief in God, and the demons were trying to deceive her to make her lose her faith. Teresa’s book, The Interior Castle,

How do you know that right now you’re not being manipulated by an evil demon? It seems that you can’t. Maybe there’s some suggestion of the evil demon’s handiwork—the fact that you’re now reading about evil demons, for example; evil demons with a sense of humor might enjoy causing people to think about evil demons. Even without hints like this, it seems impossible to exclude the evil-demon hypothesis entirely. But if you can’t know that you’re not being manipulated by an evil