One of the topics, which fascinates many geeks (like you and me) is the idea, that we all might live in a giant matrix-like simulation. And if you argue it from a mathematical standpoint, that is quite likely…but that should not be the topic this week.
This week’s paper assumes that the theory is correct, and that we indeed live in a simulation. And from that perspective it discusses different ways, how we might be able to break out of it.
At first sight, this seems quite quirky, fun and zero percent relevant to our actual lives. WHY do I put it into the newsletter? Because the ideas in the paper have implications for another field: AI safety
If we assume we build an AI in the near future that is as intelligent (or more) than humans, we might find ourselves in the very same situation:
- We are simulating a virtual environment for the AI to study it (we are the simulators)
- The AI figures out it is not in the real world, but in a simulation
- The AI tries to break out of the simulation, with the very same tactics which are presented in the paper
If you let this thoughts sink in, they are STILL CRAZY. But it is a really fun paper to read about, and you learn something new. That’s all this newsletter is about. Have fun.
Software exists to create business value
I am Simon Frey, the author of the Weekly CS Paper Newsletter. And I have great news: You can work with me
As CTO as a Service, I will help you choose the right technology for your company, build up your team and be a deeply technical sparring partner for your product development strategy.
Checkout my website simon-frey.com to learn more or directly contact me via the button below.
Many researchers have conjectured that the humankind is simulated along with the rest of the physical universe – a Simulation Hypothesis. In this paper, we do not evaluate evidence for or against such claim, but instead ask a computer science question, namely: Can we hack the simulation? More formally the question could be phrased as: Could generally intelligent agents placed in virtual environments find a way to jailbreak out of them. Given that the state-of-the-art literature on AI containment answers in the affirmative (AI is uncontainable in the long-term), we conclude that it should be possible to escape from the simulation, at least with the help of superintelligent AI. By contraposition, if escape from the simulation is not possible, containment of AI should be, an important theoretical result for AI safety research. Finally, the paper surveys and proposes ideas for hacking the simulation and analyzes ethical and philosophical issues of such an undertaking.
Download Link:
Additional Links:
- Is Reality Real? The Simulation Argument YouTube video about the Simulation theory