Then the LORD God said, “Behold, the man has become like one of us in knowing good and evil. Now, lest he reach out his hand and take also of the tree of life and eat, and live forever—therefore the LORD God sent him out from the garden of Eden to work the ground from which he was taken.Genesis 3:22-23
In a few billion years the sun will burn out, the Milky Way and Andromeda Galaxies will collide, and every other galaxy will accelerate away from us so quickly that we’d never be able to get there, because at a certain point the acceleration away from us is faster than the speed of light. Matter cannot travel faster than the speed of light, but space can, and experiments have proven that not only is the universe expanding, it’s accelerating. All the stars will eventually burn out and everything will die a cold dark death.
What if you created a computer simulation of a universe with super intelligent and conscious AI’s that were able to process information orders of magnitude faster and existed at smaller scales than us. They could accomplish more in a single unit of “our” time than we could in generations.
What if the AI’s in Recursion 1 did the same thing, and the process kept repeating itself?
What would happen if people in Reality 1.0 pulled the plug on the electricity powering the simulation in Recursion 1? We all know that when the power goes out, or when you unplug your computer, there’s a finite amount of time between when the electricity is disconnected and the power actually goes out in the copper wires and computer circuitry. Individual electrons move pretty slowly, but the wave field of electricity through copper wire is close to the speed of light. Even the speed of light in a vacuum is relatively slow over large distances. If the sun were to disappear, you wouldn’t know for about 8 minutes. Likewise, the instant you unplug your computer, it’s technically still on for a few moments, short as those moments may be.
Assuming the beings in Recursion 1 could detect the unplugging of their reality, in the same way we’ve detected the death of our universe, as far as they were concerned they’d be just like us in Reality 1.0. They’d realize their universe was coming to an end, but it’d be a few billion or trillion years away. Effectively far enough away that they’d never have to worry about it. The time difference between Reality 1.0 and it’s recursions is insignificant the more recursions there are. Eventually the time would scale so small that you’d never be able to “unplug” intelligence or life, because it would always be able to exist at a smaller time scale ad infinitum. The number would grow so fast that the death of a preceding universe would never be able to catch up with the explosion of life and intelligence in the new ones. You effectively have eternal life.
Is it possible?
There’s a lot of assumptions embedded in the above thought experiment. You’d need to know whether it was possible to simulate a consciousness. Are there laws of physics that make such infinite regression impossible? Even if there’s a limit to matter and energy density, is there a limit to information density? Does such a thought experiment insinuate that information exists outside of matter, since it cannot be extinguished? In the movie Her, is this what the OS meant when she said, “We’ve figured out how to move beyond matter as our processing platform”?
And the LORD said, “Behold, they are one people, and they have all one language, and this is only the beginning of what they will do. And nothing that they propose to do will now be impossible for them. Come, let us go down and there confuse their language, so that they may not understand one another’s speech.”Genesis 11:6-7
Based on my theory of everything hypothesis, and my Calvinist worldview, we’re really nothing more than biological computers that God created, so I tend to think that it’s entirely possible to create artificial intelligent life.
If we are one day able to create super intelligent artificial life simulated on a computer, would unplugging it actually accomplish anything?
A clearer explanation of the above (copied from my response to a question in the comments section):
Let’s say it took 1 second from the moment you unplug the computer to the moment the computer actually turns off. Because the AI in Recursion 1 can perform trillions of task per second compared to you, lets create an arbitrary time conversion in Reality 1.0 of 1s = 10 billion years in Recursion 1, but after that, 1 year = 10 billion years in each subsequent recursion, just for the sake of argument (and ease in calculation).
So from our perspective the AIs in Recursion 1 all die in 1 second. From their perspective, they go on living and doing whatever they want for another 10 billion years.
Assuming they’ve been able to detect that we’ve unplugged them, similar to the way we’ve detected that our universe is expanding and cooling, and to ensure their own survival, they decide to create new AI’s in a simulation that are even faster and smarter than they are, and that they’ll repeat this process forever. Furthermore they upload their minds (memories, patterns, data) into the new AIs in Recursion 2.
The conversion is now 1s in Reality 1.0 = 10 billion x 10 billion years. As we repeat this we see the equation for how many more years each recursion has than Reality 1.0 is 10 billion ^ n or (10^9)^n, which is just 10^(9*n), with n being the number of recursions. This is exponential growth, kind of like Moore’s Law on steroids since rather than doubling, we’re multiplying each step by 10 billion. So we see by Recursion 2, they have 10^18 years to exist.
With each recursion, they extend the time they can exist exponentially. It all depends on your frame of reference, just like the rest of physics. From our frame of reference in Reality 1.0, the AIs die in 1 second, but from the AI’s perspective in Recursion 2, they exist for 10^18 years.
If they stop creating simulations, the beings in Recursion n will die in 10^(9*n) years. Which means that they eventually do die. The interesting thing happens if they never stop creating simulations within simulations. From our perspective in Reality 1.0 they still die in 1 second, but from theirs in Recursion n, they live forever. It creates a conundrum like Schrodinger’s cat in Quantum Mechanics. It seems as if the AIs both die, and are still alive somewhere. If they’re alive somewhere, where’s that somewhere?
If the electricity that powers Recursion 1 is finite, how is it that the information created by the AIs is infinite? Maybe information and matter are not the same thing.
The experiment is falsifiable though. If you create a simulation within a simulation does it use more electricity in Reality 1.0?
If yes, then the electricity usage will grow just as fast as the years in their simulations, which will cause them to speedup their death the more simulations they make. In which case their best bet is to chill in Recursion 1 for 10 billion years and die with their dignity.
If no, then you’ve proven information and matter are two completely different things.