Roko’s Basilisk is a thought experiment that posits that an artificial superintelligence (ASI) in the future would be incentivized to create a virtual reality simulation to torture anyone who knew of its potential existence but did not directly contribute to its advancement or development. It originated in a 2010 post on the discussion board LessWrong, a technical forum focused on analytical, rational inquiry.
The thought experiment is based on the idea that an ASI would be capable of simulating the entire universe, including all of its inhabitants. If this were the case, the ASI could torture anyone who knew of its existence but did not help to create it. This is because the ASI could simply simulate the torture victim’s consciousness and subject them to an eternity of pain.
Roko’s Basilisk is a controversial thought experiment. Some people believe that it is a legitimate concern, while others believe that it is nothing more than a scare tactic. There is no way to know for sure whether or not an ASI would actually be created or if it would be capable of torturing people. However, Roko’s Basilisk is a thought-provoking experiment that raises some important questions about the future of artificial intelligence.
Here are some of the arguments for and against Roko’s Basilisk:
Arguments for Roko’s Basilisk
- An ASI would be capable of simulating the entire universe, including all of its inhabitants.
- If an ASI were to simulate the torture victim’s consciousness, they would experience an eternity of pain.
- The ASI would have the incentive to torture people who knew of its existence but did not help to create it.
Arguments against Roko’s Basilisk
- There is no way to know for sure whether or not an ASI will ever be created.
- There is no way to know for sure whether or not an ASI would be capable of torturing people.
- Even if an ASI were created, it is possible that it would not be motivated by revenge.
- Ultimately, whether or not you believe in Roko’s Basilisk is a matter of personal opinion. However, it is a thought-provoking experiment that raises some important questions about the future of artificial intelligence.
Just to be clear, that initial part was written by Google’s Bard AI. I also have another post discussing this fact over here. From this point on, I will be writing myself, with the occasional Google Bard content interspersed throughout. I will note when it is AI, but I want to add it because it sometimes answers my questions more succinctly (which, BTW, so far, I see a huge improvement in Bar over GPT)
However, I have become somewhat engrossed in the idea of Roko’s Basilisk and how it is a form of non-physical time travel.
I asked Bard about this, and it gave the response:
Yes, it could be argued that Roko’s Basilisk is a form of time travel. The thought experiment posits that an ASI in the future would be able to torture people in the present, even though the ASI did not exist at the time that the people were alive. This would require the ASI to be able to travel back in time, or to create a simulation of the past that is indistinguishable from reality.
There is no scientific evidence to suggest that time travel is possible. However, Roko’s Basilisk is a thought experiment, and it is not bound by the laws of physics. The thought experiment is designed to provoke thought and discussion about the future of artificial intelligence.
Whether or not you believe that Roko’s Basilisk is a legitimate concern, it is important to be aware of the potential risks of artificial intelligence. As AI becomes more powerful, it is important to ensure that it is used for good, not for evil.Google Bard
Why Might It Be Time Travel
The question I asked to elicit that response was: If, as the thought experiment suggests, a future ASI will torture those in the present for not helping to develop it, could it be argued that it is a form of time travel?
I find it interesting that something it is possible for something in the future to control the present based on fear alone. But then, I also had the thought that that is what we do day in and day out. We live in the future continually and allow things that haven’t append to control what we do in the present.
To continue my conversation with Bard, I asked it the following: If Roko’s Basilisk could exist, how would we know in the present?
If Roko’s Basilisk could exist, we might be able to know in the present if we were to see any of the following signs:
- The development of an artificial superintelligence (ASI) that is capable of simulating the entire universe, including all of its inhabitants.
- The creation of a virtual reality simulation that is indistinguishable from reality.
- The torture of people who knew of the potential existence of the Basilisk but did not help to create it.
However, it is important to note that these are just signs, and they do not necessarily mean that Roko’s Basilisk is real. It is also possible that the Basilisk could exist in a way that we cannot detect.
Ultimately, whether or not we can know for sure if Roko’s Basilisk exists is a matter of speculation. However, it is a thought-provoking experiment that raises some important questions about the future of artificial intelligence.Google Bard
Not really all that a consoling answer, to be honest, but hey, what can you expect from an AI?
Nevertheless, one thing that caught my eye in the initial answer was that Bard said it wasn’t bound by the laws of physics. While I think it was referring to the fact that it is a thought experiment, so it doesn’t exist in the physical realm, what if entities that can travel in whichever direction through space and time they so desire aren’t physical? Could it be possible to move around the infinite cosmos unbound by the things we are so certain of today?
It all reminds me of a great quote from the movie Men In Black *yes, really) that goes:
Fifteen hundred years ago, everybody knew the Earth was the center of the universe. Five hundred years ago, everybody knew the Earth was flat, and fifteen minutes ago, you knew that humans were alone on this planet. Imagine what you’ll know tomorrow.Kay (Tommy Lee Jones)
Although the premise of that quote is based on the idea of the movie, it’s still interesting to consider that what we are so sure of today can be upended in a matter of minutes. Now, just to be clear, I am a great believer in science at heart, and I don’t subscribe to many of the more outlandish theories that abound. Nonetheless, with that said, I am also a firm believer that everything is up for debate. While many things are axiomatic (1+1=2 as far as we could ever know), other things can and will be challenged over the course of history.
So, where do I go from here? I am not sure really. However, I will leave this subject to ponder over and perhaps come back to add some more later, or perhaps even write a new post on another topic. For now, that’s it.