Why don't computers think like humans?
We almost certainly live in a computer simulation
George Hotz is one of the most famous hackers in the world. He came to fame and honor in 2007 when he - at the tender age of just 17 - cracked the iPhone with a "jailbreak"; Nobody had done it before him. Today he works in his company Comma.ai on open source software for self-driving vehicles.
In March 2019, Hotz gave a talk at the South by Southwest creative conference in Austin, Texas. The audience would have expected it to be about self-driving cars or being a hacker or something like that. Far from it: Hotz explained to the (presumably slightly confused) audience that he was convinced that we all lived in a computer simulation - and that he made it his goal in life to find a "jailbreak" for our simulated reality.
Hotz is neither a joker who has allowed himself a joke, nor is he mentally deranged. The idea that our reality is a computer simulation is actually a serious philosophical problem. And one with serious consequences.
The simulation hypothesis
When we hear that our reality could be a computer simulation, most of us think of The Matrix, the famous sci-fi trilogy. The scenario of "The Matrix" is (Spoiler Alert!) That humanity has been subjugated by artificial intelligence and we are in a kind of vegetative state: the perceived reality is a computer simulation, and in fact we are slaves to the AI.
The scenario from "The Matrix" is not the philosophical simulation hypothesis. This goes even further: We are not trapped in a computer simulation - we are the computer simulation. We only exist in simulation. We don't exist in real reality; everything we are and everything that surrounds us is completely simulated.
The most famous variant of the simulation hypothesis was proposed by the philosopher Nick Bostrom in 2003. Bostrom argues that it is conceivable, and perhaps even highly probable, that we are a man-made computer simulation. As soon as the computing power of real mankind became large enough, it is conceivable that people began to run simulations of their past. In terms of logic, it would be like trying to simulate ancient Rome today to better understand how the ancient Romans lived.
The simulation hypothesis that George Hotz and Elon Musk believe in is more general. Something like humanity must never have existed in real reality; it is sufficient if some civilization or intelligence has simulated one or more universes in real reality.
So far, so interesting - but the simulation hypothesis doesn't sound like much more than one episode of "Black Mirror". But why exactly do smart people like Hotz, Bostrom and Musk think that the simulation hypothesis should be taken seriously? It's all a matter of probability.
A game of probabilities
If we compare modern video games with video games from 30 years ago, the contrast is as strong as day and night. Today we can experience three-dimensional virtual worlds on any PC or smartphone in a wealth of detail and complexity that was practically unimaginable a few decades ago. This development may not go on in this linear way forever, but we can assume that in another 30 years huge leaps will be made, and so on.
At some point our computing power and our programming skills could reach such a high level that we can create simulations with an unimaginably high level of complexity today. If we ever reach this level of technological maturity, then it is almost certain that we will do the complex simulations that we can run.
One simulation that should be very interesting for us is to simulate the universe fresh from the Big Bang, for example to better understand the laws of nature. In order to collect data that is as usable as possible, we don't just start a single simulation, but maybe a thousand. Or ten thousand. Or a million. Or a hundred million.
In the simulated universes, simulated creatures evolve, then simulated primitive civilizations, and then finally technologically highly developed civilizations. For their part, many of these simulated civilizations eventually develop enough simulated computer knowledge to run their own universe simulations. Our future universe simulation would have the consequence, with a probability bordering on certainty, that millions or billions or orders of magnitude more simulated civilizations will emerge.
And that is exactly the crux of the simulation hypothesis. If we assume that at some point it will be possible for us to simulate something like universes, then there will be innumerable simulated civilizations. From a purely statistical point of view, however, the probability that we really are the one special civilization in real reality that sets the simulation stone rolling is practically zero. Because what we will observe millions or billions of times in our future simulations (civilizations that evolve to the point that they can simulate universes) is identical to what we will also observe for our own civilization (a civilization that has evolved to the point where it can simulate universes).
Which speaks against the simulation hypothesis
The simulation hypothesis is almost frighteningly elegant in its simplicity: the probabilistic argument is clear and basically watertight. Does that mean we are definitely living in a simulation? It's not that simple.
The simulation hypothesis is difficult to handle from an epistemological point of view. On the one hand, the simulation hypothesis is basically plausible, in contrast to religious creation myths, for example. On the other hand, there is no empirical or other evidence to suggest that the simulation hypothesis could be true. So the evidence is completely absent. In addition, the simulation hypothesis is not really refutable, which at least to some extent raises the question of whether the simulation hypothesis is more than an amusing metaphysical mindfuck.
Another uncertainty surrounding the simulation hypothesis concerns our future computing power and programming skills. We can assume that our computer simulations will get better and better in the future. However, it is unclear whether more computing power and better programming skills are sufficient to make certain fundamental factors of our reality simulable.
One such factor is sentience. In the 1970s, the philosopher Thomas Nagel asked what it actually felt like to be a bat. His argument is that we can understand (or simulate) a bat (or any other sentient being) down to the last atom, but that this alone is not necessarily enough to produce the bat's sensations or "qualia".
Today, Super Mario may romp around in extremely detailed and large worlds, but when Bowser heats him up with his fiery breath, Mario feels as much as he did 30 years ago: absolutely nothing.
Maybe we're living in a simulation. So what?
We either live in a simulation or we live in real reality. Is this question even relevant? Even if we live in a simulation, it does not relieve us of the daily grind; we still have to get up tomorrow morning and go to work so that we can pay our bills at the end of the month.
However, the simulation hypothesis is not entirely without reference to everyday life. Should there ever be more concrete evidence that we are living in a simulation, for example, all religions should become obsolete because they would be definitively refuted. In general, our view of ourselves and what our society should be could change. If we knew that we are all being simulated in action and truth, we might rather pull ourselves together and try to resolve our conflicts. Does it really make sense for two simulated states to fight for a piece of simulated land?
Perhaps the most important reason to deal with the simulation hypothesis was given by George Hotz in his talk at South by Southwest: If we live in a simulation, we can try to hack the simulation. So we could perhaps make our reality a little less unfortunately filled and brutal (one or the other cheat code would be quite helpful to make our lives more livable). And maybe we could even catch a glimpse of what we can only speculate about at the moment: Real reality outside of the simulation - and our actual creators.
This article appeared for the first time on "Watson".
- Why is Laravel good for web development
- Are you happy in a love marriage
- What government jobs are there in retail
- Faster web surfing with many tabs open
- How do I make a killer mixtape
- Excel Pivot Tables 101
- How much auto insurance in Alberta
- What is temperament in psychology
- How will black Americans be suppressed in 2019?
- What does it mean to be drunk on this love
- Who is the inventor of Facebook
- How can you kill laziness while programming
- Why is Zayn Malik so cute
- Will aluminum foil block 5G
- Which philosophers do you most agree with
- What do you know about grass
- What are the best cameras under 20K
- What is ellagic acid
- What is cytosol Where is it located
- Are hybrids harmful to health
- What does the constitutional state mean
- What is the LayoutInflater in Android
- In fact, has Lowes blacklisted former employees
- I don't know what to do
- How are decoders and multiplexers used in practice?
- Where can I buy a frequency sensor
- How can you meditate
- How is secondary school life in America
- Why is China's energy consumption increasing
- What bad things bring you joy
- Why did Laura Secord die
- Why is man not happy today?
- Why do you support Hong Kong
- Does one really remember experiences for life?