Roko’s Basilisk
An interesting and terrifying thought experiment.
Disclaimer:
What you’re going to read is considered by some a very terrifying thought experiment that can be n information Hazard to some of the people. An information hazard is something that can potentially harm you or people surrounding you. So if you’re reading this just know that the choice that I made by writing this blog is that I don’t care about its content and didn’t seem that terrifying to me. So continue at your own risk.
The Beginning:
The beginning of this experiment was something of an internet legend. A few years ago, a man named Roko published a blog on the rationality and philosophical site “Lesswrong”. This blog was a thought experiment as Roko called it, when the founder of the blog, Eliezer Yudkowski, read it he called Roko and idiot and his blog stupid. However after a while the blog was deleted and prevented any further discussion about this experiment.
The Basilisk:
Roko’s thought experiment was named after a mythical beast named Basilisk. It’s said that the beast can kill any adventurous hero with its gaze only. Many of you might have heard about it in the “Harry Potter Series” or some other kind of Greek or fantasy story book and apparently Yudkowski deleted the post because he thought that the basilisk, metaphorically, cause harm to the users of the website as they gaze down their screens. So what he did was not slay a mythical beast but an idea. Many users of the website did report great amount of mental distress and nightmares after considering this idea so maybe it is true that it can cause harm. So once again I’m warning anyone who reads this to be self aware of the fact of what you’re reading.
The Experiment:
The experiment starts something like this; Suppose that in the future we make a hyper intelligent AI, something start out of the singularity and then we ask that AI to help us optimise all aspects of human civilization and then for some reason to us ignorant humans it decides that the first step towards optimization is starts with inflicting eternal torment on every single human being that didn’t want it to come to fruition or didn’t help it come into existence in the first place, after all how can you optimise without the optimizer.
How can it know? Well a being like Basilisk is so intelligent that it can simulate all of human history and every single human faster than a snap of finger. Hence it can meaningfully predict what exactly a person thinks and will think and be confident enough that you will do exactly as it predicts, in short it can read your past to predict your future.
Now comes the interesting part to this experiment, since it can know what you did or didn’t for its coming into existence then shouldn’t everyone collectively start bringing it into existence from now? So there’s 2 news, the good one is you never knew about its existence so you could escape the torment and the bad one is that since you read it now you probably have fallen under the basilisk’s radar and now you have to bring it to fruition.
The Newcomb’s Paradox:
To make the Basilisk a tad bit more sinister there’s another thought experiment known as Newcomb’s paradox and it follows as- Suppose there are two boxes A and B in box A there’s a $1000 and is transparent. In Box B, well it’s opaque so you don’t know what is in it. Now I come in and say look, I’m a great predictor and if choose box B then you get a million dollars. And if I predict that you will choose both the boxes then you’ll get only $1000 as the other box would be empty. Now what would you do? Now the answer to this question ultimately lies on you thinking how good a predictor I am, right? Enter the Basilisk, say that it presents you with a similar choice and it wants you to choose both boxes A and B though as it will metaphorically bring it to existence. And it then tells you that if you take box B then the other option for you is eternal torment with wires fitted all over your body and living a life equivalent to thousands at the same time and all of them horrific as hell(sorry, not sorry). Now you are convinced that the basilisk is a perfect predictor so naturally you will take the both the boxes, and since it knew you would take both the boxes when under tremendous pressure the other box doesn’t have to contain any eternal torment at all.
The shock:
Now think about it, you have experienced future blackmail by something that doesn’t exist yet. And if you think that the there’s even a slight possibility that the basilisk is real then you should bring it into existence right now. All of this simply because you had a single thought about it. You will have acted in its past the way it will have wanted you to act in its future.
The Conclusion:
Now to end this, all I want to say is that even if the Basilisk doesn’t exist when we think about it and the more number of people think about its more likely for it to come into existence. It’s like an idea virus and now we know why Lesswrong handled it this way. I don’t think we will achieve this hyper intelligent malevolent AI that shoot beams from its eyes and kill you. If you’re scared then let me tell you it’s highly unlikely that this type of command will be given to it and if we collectively think not to bring it into existence then it wouldn’t really come into life.
For a more interesting visual depiction of the blog you can checkout my inspiration for writing this:
