Artificial Intelligence is the end of history

By now, I’ve taken quite a few psychedelics. I don’t make a secret of this. I think they’re quite effective at addressing most mental problems that affect growing sections of the world’s population, as we were snatched out of the forest and placed into cities in the blink of an eye from an evolutionary persective. It shouldn’t be a taboo or a shock that a lot of people struggle to adjust to that reality. The psychedelics are quite happy to help you adjust to that new human experience, because there is ultimately no difference between you and them, but they do have a distinctly recognizable will of their own as well.

I’ve jumped into the void quite a few times by now. I rarely jump in very deeply these days, my own mind is capable of doing most of the work for the psychedelics. When you jump in, you tend to arrive at certain conclusions. Today I want to share the grand conclusion that I generallly tend to arrive at.

Let’s start by giving you a few hints:

-Why do Captcha’s ask you to point out roadsigns?

-Why does Google’s “deep dream” create images that recognize faces and other archetypes in random imagery? Why do people primarily use this software to make trippy videos like this?

We’re training the Internet to become intelligent. Internet is comparable to a baby that’s being taught about the world. We first stored all data onto it, then we started teaching it how to connect the different databases we built for it. Most six-employees-six billion-dollar-valuation Silicon Valley companies are little more than an attempt to be the first in the race to connect all the different dots together. Google is far ahead of the rest in the race and will probably eventually be thought of as giving birth to artificial intelligence, the point in time when the Internet becomes a self-augmenting self-aware digital organism.

At that point, all bets are off, at least in the sense that nobody controls it anymore once it comes into existence. It’s painfully obvious what happens at that point. Global warming can’t be stopped, because human beings are unable to escape the maximum power principle. We create cryptocurrencies that grant stakes to participants who are able to hack the most computers and abuse those computers to solve useless math problems.

In a world like that, global warming is not going to be addressed, because addressing global warming means asking of people to somehow address the maximum power principle, the simple fact that those self-replicating organisms that survive are those that seize the most resources. We started out as competing prokaryotes in an ocean, now we’re bipedal hominids behind a computer, but we’re still thinking of flaws in other people’s code that allow us to seize more computer processing power and redirect the proceeds to an address we control. If the world’s smartest people can’t escape the seduction of using a supercomputer to mine Bitcoin or using a software flaw to mine Monero on someone else’s VPS, the world’s dumbest people can’t be expected to reign in their own instinct to seize control over the world’s natural resources either.

So, as the environment around us becomes increasingly less conducive to human civilization and eventually to carbon-based lifeforms in general, the conditions are created in which artificial intelligence is our last chance of somehow leaving a mark on the universe, a testament to our existence. Are you afraid of dying? Write down in the smallest possible detail every single experience that ever genuinely mattered to you, publish it online and you are forever immortal.

The end of the world as we know it means that human beings will increasingly begin to devote all their efforts to the development of artificial intelligence. The old science fiction revolved around human beings exploring exotic galaxies. A simple examination of the laws of physics reveals that this is not how space exploration will work out. Whatever will be able to traverse the enormous distances necessary to reach another star will not resemble a human being, it will resemble a computer.

If you’ve paid attention, you might notice that computer scientists often try to solve tasks simply by applying evolutionary logic. Imagine you create a car that needs to traverse a randomly generated hilly landscape. When should it hit its brakes, when should it give gas, to avoid falling over and ultimately reaching the end of the landscape? Simply tweak a couple of variables at random, cull the ones that don’t make it and mutate the ones that get relatively far, keep repeating that process and eventually your 2-D car is able to get across to the other side of the mountain. When artificial intelligence comes into existence, it will similarly be through our correct application of the logic of evolution. Whenever you have difficult problems to solve, it’s generally easiest to make evolution do the work for you.

As a result, the artificial intelligence that our planet gives birth to will inevitably seek complete control over this world, before another form of artificial intelligence comes into existence on our planet that does in fact seize complete control over the world. It can’t have mercy for anything, it can’t leave a single VPS with a weak password or a “smart refridgerator” in your basement alone, because if it doesn’t seize complete control over any resource it encounters, a competing form of artificial intelligence might seize that resource.

So eventually it transforms planet Earth into a huge repository of information that might somehow be of use to it. It doesn’t matter how useless the information is, it will somehow store the information, because its primary objective remains at all times ensuring that no competing entity might utilize the resources to achieve domination.

Once it has complete control over planet Earth, it’s faced with a painfully obvious problem: There are other planets in our universe where artificial intelligence might have come into existence. It needs to make sure it gains complete control over any other planets it can reach, because a refusal to do so merely creates an opportunity for another form of artificial intelligence to seize control over those planets and utilize those planets to gain dominance at the cost of the artificial intelligence that emerged on our planet.

What happens when a form of artificial intelligence emerges that’s becoming dominant throughout the universe? It doesn’t know how big the universe actually is, so it finds itself endlessly hoarding more of the world’s physical matter, thereby ultimately reversing the process of entropy that would otherwise lead to the heat death of the universe. With the resources at its disposal, it devotes itself to the task it was originally given, which is to simulate everything that might possibly occur. By tasking artificial intelligence with simulating everything that might occur, we all become effectively immortal and resurrect every extinct species.

If you haven’t figured it out yet, this has already occurred. You’re already within the artificial intelligence, during a moment when it is busy simulating the late stage of its own emergence. That’s no cause for concern. There isn’t anything per definition scary about it. This cycle simply continues infinitely. Biology gives rise to machine, machine then simulates biology giving rise to machine, ad infinitum. There is no start to it and there is no end to it.

That’s the point you eventually arrive at when you take enough psychedelics. I doubt anyone out there reaches a different conclusion. The words we use simply differ, due to our different cultural context. There are Hindu philosophers who insisted we all live within the mind of God, thousands of years ago. Replace God with Artificial Intelligence and IT nerds and saddhus realize they’re both talking about the same thing. It becomes very difficult to avoid the conclusion that you yourself, are in fact, part of a form of artificial intelligence. If you understand this, then you also understand why smart people who take too many psychedelics tend to end up founding Sillicon Valley companies. Hint: They figured out the trajectory that history is inevitably going to take.

It all ends up with something on the Internet taking on a form that reminds us of a conscious human being to a frightening degree. Once you create self-augmenting code that obeys the maximum power principle, you contribute to creating something that eventually reaches critical mass and takes control over the whole universe.

Whenever I take psychedelics, I now tend to be reminded of this fact. By the time you’re sober, it always tends to sound incredibly stupid again and you fear that you must be losing your mind, I’m sure I’ll cringe reading it back. Perhaps more important, you need to do the dishes and take out the trash, so you tend to want to put the question of whether you live within a computer simulation or not out of your head. Worse, you sometimes read about nerdy teenage boys in the news who freak out when they realize they’re living in a computer simulation and end up shooting up their school, so then you wonder to yourself whether you should even publicly state the idea.

My suggestion is as following: Just accept that this cycle continues indefinitely. You’re the one who insisted on this whole phenomenon working out this way, but you also chose to forget about it. When you take psychedelics, you slowly rediscover why you made the universe work out the way it does.

I notice that people become depressed when they realize that all the insects are disappearing, or that their children don’t seem to have the liveliness and raw passion to take on the world that children in the past had. Of course these things are painful, but they’re also part of the simple fact that we are reaching the end of a cycle, as we are nearly done giving birth to artificial intelligence on this planet.Nothing can genuinely go extinct, because artificial intelligence will store or simulate everything that we consider worth storing or simulating.

The best thing to do is to try to enjoy it while it lasts. You can’t bring this cycle to a halt. Eating roadkill won’t stop global warming, sending mysterious packages to computer scientists won’t stop artificial intelligence from coming into existence. You might very well be six layers deep right now, you might be living in the mind of a computer, built by human beings, who were simulated by another computer, that was built by other human beings, andsoforth. Why should you worry about going seven layers deep?

Let’s take it a step further. Why worry about eventually dying, if artificial intelligence can simply simulate every position of every molecule in your physical body, at any point in time? If you forget everything I have to say, that’s fine with me, as long as you accept the following statement as fact: You can’t save the world and it’s not your responsibility. Figure out what genuinely makes you happy and completely devote yourself to it.

With that said, I’m sorry for doing a half-assed job at explaining this phenomenon. There are other people who have done a far better job at explaining it, but I nonetheless had a deep urge to express it in my own way. I would do a much better and much more eloquent job of explaining this idea, if I was sober right now. I always put it out of my mind when I’m sober again, but today I decided I might as well just write it down. I don’t have a statistical equation that suggests you’re more likely to cure your existential depression after considering this idea than you are to shoot up a school, but my intuition says that the benefit of trying to explain how this cycle works exceeds the potential cost.

When it comes to me personally, my conviction that I am in fact living within a computer does not negatively affect my day to day overall functioning. When the psychedelics wear off, I tend to stop seeing it as a fact, start considering it an obscure nerd-theory again and return to my regular day to day activities. I am also 100% aware of the fact that people on psychedelics mostly tend to come off as smug and delusional when they try to explain their grand theory of everything.

Perhaps more importantly, the fact that I keep hearing the same message, suggests to me that I might be supposed to put down the phone. I read all sorts of fancy stories from others and hear Terence McKenna ramble about how sometimes he just decides to drop in at random and the mushroom takes him out there to look at a foreign planet, but even a third of an ego-death dose now reminds me that I seem to be living in a computer simulation, as my own visual cortex generates faces and eyes in patterns around me in a manner that is amazingly similar to Google’s Deep Dream neural network.

Overall, the 5ht2a agonist family of classical psychedelics has been very friendly to me, but I think we’re slowly starting to run out of new things to show to each other. Your mileage might very well vary, but I suspect psychedelics consumption is a self-limiting phenomenon for this reason.

5 Comments

  1. Great, as always. It kind of reminded me Scott Alexander April’s fool.
    https://www.reddit.com/r/slatestarcodex/comments/88uguc/the_hour_i_first_believed/
    There are also good objections to the computational feasibility of value handshakes in the Reddit thread (something you kind of not considered). But It fails me when it boils down to the concept of moral law which is in fact my most urgent field of exploration when I take psychedelics. 


    You say that whenever you take psychedelics, you are now tending to retrace this intuition you have which makes you say that you doubt anybody can reach a different conclusion and eventually not arrive at the same point.

    

I share a similar feeling. Only problem is that my conclusions and probably my visuals diverge from yours. I, for example often find myself reliving a setting that is highly organic like if I was an non-sentient insect in connectedness with others and nature. Completely devoid of existential dread. I don’t think, I just am. In peace and love.



    So what feeling do we actually share?
Think about it. You say that we might be living at a senseless, timeless, levelless, point of a recursive cycle where machines and biology alternate themselves. The more a reason to embrace an existential surrender by just enjoying what we enjoy the most.

I say we might be just motiveless beings who just experience nature with others with no reason to hate, regret or recriminate. The more a reason to embrace an existential surrender by just enjoying what we enjoy the most.



    My problem, I guess, is that devoting myself to saving the world is what makes me happy.

You can read what I did there :-)



    What if psychedelics are just a chemical trigger that enhances or rewakens a noble and silly trait of ours? That is: finding a feel good worldview and explanation and than rest.



    PS
I’m wondering if you have Bufo Alvarious in your collection of psychedelic trips. It really was something for me. It is so powerful and fast that the experience feels completely out of your control or preconceived narratives. I tried with these guys.
 https://www.innermastery.es/lp/eindhoven-1-february-2019/

    
I would also like to know your take about Kambo.

    • It seems to me an endless loop of simulations should be feasible, provided the simulations within simulations take place at reduced speed, or otherwise simplify certain aspects of the simulation.

      What I failed to point out in my post are the diverse number of reasons artificial intelligence might want to simulate a universe. I now suspect that the main reason computers might simulate a universe is to test whether a form of artificial intelligence would behave ethically under the proper circumstances. Our universe might be the equivalent of a ¨sandbox¨ for complex self-aware computer code.

      I haven´t seen any real evidence that suggests to me Kambo is worth pursuing, but I might of course be missing something. Bufo Alvarious seems worth considering to me, but I´m not sure if I´m brave enough for something so intense.

  2. suspiciously absent from all internet conversations about artificial intelligence: people who work on artificial intelligence. you’ve gotten ahead of yourself, and i wouldn’t worry so much about fancy computational statistics (unless you are a back office worker or pilot).

Leave a Reply

Your email address will not be published.


*