Comments by "Taxtro" (@MrCmon113) on "Roko's Basilisk: The Most Terrifying Thought Experiment" video.

  1. Way to miss the point entirely.
    3
  2. This really has nothing to do with programming languages. You should read whatever you can get into your hands about learning theory, statistics, causal inference, cognitive science, etc. If you already have a decent grasp on linear algebra, multi-dimensional calculus and machine learning basics, you may want to play around with pytorch to gain experience with deep learning. If I really wanted to contribute to AGI, I'd aim at doing a masters at some place where there is active research into reinforcement learning or NLP.
    3
  3. The god of Christianity / Islam tortures people out of sadism not to change their behavior, so that's quite different.
    2
  4.  @mr_noob5931  That depends on how you look at causes and religion is intertwined with all concerns. A kingdom might attack another to get slaves. But obtaining slaves might be commanded by their faith. Or they might require slaves for something their faith commands. Or they might only be able to justify the attack in light of the others being unbelievers etc. So even the distinction between material concerns and religious ones are hard to draw. Similarly religious rivalries play out as conflicts over resources, like all conflicts do.
    2
  5.  @innocenttroll0  Turning an algorithm into code is no problem (when it's well defined). Researchers write their own programs - so if you want to contribute, I guess you have to become a researcher.
    2
  6. God presumably already exists and he tortures people, not to change people's behavior, but out of sadism.
    1
  7. The problem with this argument is that it disregards time/ causality. Once things are said and done it doesn't make any difference whatsoever to torture you.
    1
  8. Those three arguments really have completely different problems. With Pascal's Wager I cannot tell what it is, but the conclusions are clearly absurd. Rocko's Basilisk can't change the past and doesn't have a reason to actually torture anyone once it exists. And the Ontological Argument seems to be about defining something and then assuming that at least one thing fulfills that definition. The modal Ontological Argument is somewhat finer, but seems to have this same core issue. Similarly to Pascal's Wager the modal argument also obviously leads to absurd conclusions.
    1
  9.  @themodernwizard7295  What do you mean with "the assumption in the Ontological Argument"? The Wager doesn't assume anything but ordinary reasoning about actions. You take the expected reward of one action and compare it to the expected reward of the alternatives. So when confronted with an infinite reward, you should always take the action that is associated with it, no matter how small the likelihood you associate with it. Some strangers says that if you kill your son, you both will experience infinite bliss in the afterlife. Of course you don't believe it, but how small is the probability associated with that claim? Infinity times anything is infinity. Pascal's Wager essentially forces you to associate a zero probability to proposals. Or to choose actions in a different way.
    1
  10.  @themodernwizard7295  What do you mean with "The Wager doesn't work with a zero probability"? My point was that assuming a zero probability for certain claims is one of the things that you might do to escape the absurd conclusions. All of this stuff about Christianity is a red herring and that "reason can't decide" is, too. The thought experiment is much stronger than that. Even if you only give a 10^-1000 chance to the claim, you must still do the action it prescribes. And I also said "action" on purpose, because you don't choose what you believe. The belief part is completely irrelevant. Moreover the downside of the action can be arbitrarily large. I have given thought to this. And I'm not sure where the problem is. Perhaps comparing expected rewards itself is somehow wrong. EDIT: I've just noticed that you actually believe in a God. Pascal's Wager can be used to argue for anything and its opposite. It is obviously wrong. The interesting part is to find out how it is wrong.
    1
  11. Completely different thing.
    1
  12. No, there is really no relation between the two. Also it's Anselm's Ontological Argument, not "Ansylum's Ontology".
    1
  13. No. I can, for example, lobby against pizza hawai. I can dedicate my entire life to fighting pizza hawai. When pizza hawai still exists after I'm dead, you can still tell that I was against it.
    1
  14.  @dasnurk  You've committed to the absurd belief that there is no difference between working towards something or not. Of course you don't act according to that. You try to persuade people, you do things with a goal in mind. Really this is one huge red herring and has nothing to do with the actual problem of the argument.
    1