Hacking the simulation

According to this piece, two billionaires (Elon Musk and Peter Thiel?) want to help break humanity out of a giant computer simulation.
"Many people in Silicon Valley have become obsessed with the simulation  hypothesis, the argument that what we experience as reality is in fact fabricated in a computer; two tech billionaires have gone so far as to secretly engage scientists to work on breaking us out of the simulation."
In this post, I indulge this view, and in the spirit of April 1st, I go with it.

Hacking the simulation by breaking its multitenancy and overloading the VMs

What do the aliens computers look like? Do they have racks and datacenters? Or are their computers more at the molecular/analog level, and avoid the inefficiency of overlaying a digital layer over the physical one. Quantum bits seem so bizarre that I think it follows directly from an elemental/basic construct of the aliens' simulation hardware.

Anyways... Since these simulations are very detailed, and even simulations of branching multi-verses, they should have distributed clusters dedicated to the simulation. And if they are worth their salt, they should be using multitenancy to prevent underutilization of hardware and reduce costs of these heavy simulations.

Is this simulation at the person level? I think it is at least at that level if not at a more finer level. Anything larger would be too coarse. So let's bet on the person level.

If this is a distributed multitenancy deployment simulation, many of us (i.e., many people) share the same machine for the simulation. And we are overpacked based on the statistical likelihood that when some of us are using more resources some others use less resources. This way we co-inhabit the same machine in a crowded manner without any problem. This multitenancy setup helps keep computation costs low, a fraction of what it would be without using multitenancy.

Based on these assumptions, we can try to trick the multitenancy system in order to overload some  machines. The trick is to first do nothing, and let the load-balancing system pack way too many of us together in the machines. If, say, 100 million of us do nothing (maybe by closing our eyes and meditating and thinking nothing), then the forecasting load-balancing algorithms will pack more and more of us in the same machine. The next step is, then, for all of us to get very active very quickly (doing something that requires intense processing and I/O) all at the same time. This has a chance to overload some machines, making them run short of resources, being unable to meet the computation/communication needed for the simulation. Upon being overloaded, some basic checks will start to be dropped, and the system will be open for exploitation in this period.

*wink, wink* Maybe this is why @elonmusk invest his time on its Twitter account to reach to many millions of followers. He keeps posting memes, rap singles, and controversial statements and keeps increasing his follower count. Maybe one day he can convince his followers and coordinate them to give the above scheme a try.

How do we exploit the system in this vulnerable window?

In this vulnerable window, we can try to exploit the concurrency cornercases. The system may not be able to perform all those checks in an overloaded state.

Maybe then we should try to find cornercases like Douglas Adams's description of flying: You jump forward, and then while going down you get distracted by something, and miss the ground accidentally, and that is how you fly.

We can also try to break causality. Maybe by catching a ball before someone throws it to you. Or we can try to attack this by playing with the timing, trying to make things asynchronous. Time is already a little funny in our universe with the special relativity theory, and maybe in this vulnerable period, we can stretch these differences further to break things, or buy a lot of time.

What are other ways to hack the system in this vulnerable window? Can we hack the simulation by performing a buffer overflow? But where are the integers, floats in this simulation? What are the data types? How can we create a typecast error, or integer overflow?

Can we hack by fuzzing the input? Like by looking at things funny. By talking to the birds or jumping into the walls to confuse them.

What are other possible ways? I am not well versed with cracking/exploiting computer systems. Come on, we should all do our parts.

Comments

Anonymous said…
Clearly there’s something about DMT that affects how we look at reality. The brain produces as much DMT as serotonin in the visual cortex and the visual cortex is responsible for how we view reality. It may be possible to alter the blueprints to reality through mind altered states of consciousness. One theory about reality is that it actually exists simultaneously in many different forms and there are mechanisms causing the brain to only see one version of reality and that dmt or something lifts the veil so to speak to allow us to perceive infinitely more information. If so, we may be able to use our minds to interact with, interface with and control it.

I believe the quantum state is how the reality can be simplified and approximated.

The double slit experiment shows that when we aren’t looking at reality, it is algorithmic. When we are, it’s material. Dream/astral state or hallucinations may be a view of algorithmic reality. The reason we have prophetic dreams is as much about the link between the future and the preset through the quantum state as it is about the visual manifesting reality through subconscious causality.

The mind doesn’t distinguish between real and imagined reality. It also seeks self consistency. So if you give it a reality that is different and inconsistent, it will try to make that vision more consistent by actually acting to make it happen. The reason the Simspsons “predict” reality is they are creating it. The unusual drawings of the characters sticks out in our minds as different and the brain can’t deal with the inconsistency so when they have Trump become president in an episode in the 90s or other spooky predictions, we subconsciously act in ways to materialize it.

The past may also be invented as needed so we may not have anything like a consistent past. This explains why eyewitness accounts don’t work. People invent the past. Algorithms might approximate it based on available information, so if we can alter the information available, we may be able to rewrite our past.

This can be done for a variety of purposes. In one sense, anyone who’s traumatic memories of the past affected them can manipulate it through hypnosis so it no longer affects them the same way. They can literally give themselves a different past that will change them for the better... In another sense, people can build a story ark of cause and effect that imply direction and give themselves a new backstory. So they give themselves a story where they were always picked on even if they weren’t, and how they outsmarted and overcame and learned valuable lessons that shape who they are today. The story we tell ourselves is very powerful narrative ark that determines how we live.

Alternatively, we can leave a message for the future, that implies a different past for them or near future for us and influence the future… you could look at the current trajectory of space travel and if you determine that this is ultimately going to shape humanity, you make a tshirt about being a part of a Martian colony in July 2028 to accelerate the timeframe. Presumably if you randomly stash 10 shirts throughout different parts of the world or strategic locations like give them away at an actual Space X facility, the distant future will eventually lose the story of how those TShirts came to exist and they will have to give their past a story which means they have to hold a story that in July 2028 there wa an effort to colonize Mars.

So the theory is that you influence our near future to occur. Reality must adjust to create a self consistent time line. The future influences the past through retrocausality.

Popular posts from this blog

Hints for Distributed Systems Design

Learning about distributed systems: where to start?

Making database systems usable

Looming Liability Machines (LLMs)

Foundational distributed systems papers

Advice to the young

Linearizability: A Correctness Condition for Concurrent Objects

Understanding the Performance Implications of Storage-Disaggregated Databases

Scalable OLTP in the Cloud: What’s the BIG DEAL?

Designing Data Intensive Applications (DDIA) Book