Proposal to Brute Force Artificial Intelligence

We live in a world where there are only two possibilities; exist or not exist — 1 or 0. An approach to solve intelligence could be through repeated continuous simulation of 1 and 0 where we start with simply 1 or 0 where we will be able to start with concepts such as 1 and 0 where the branch that lives will constantly reproduce. Then, each 1 and 0 will branch again and whichever one does not move on will produce an ‘object’ or ‘concept’. Using the example of starting with 1 or 0: when we branch off, the branch that is randomly assigned 1 will move on and continue to evolve. The number that guesses 0 will get it wrong and the ‘0th’ concept will be formed which could possibly be the very idea of 0. The failure to move on will cause the existential code to be added to the environment. The proposed environment will allow any being to view any other concept in this environment, but only a sufficiently complex existence or nonexistence will be able to use and understand the environment around it. Moving on and branching off of the 0th concept, we will get a 00 concept and a 01 concept. 00 will start to become more of a more complex concept which will continue to evolve as it continuously guesses 0. Of course, this will constantly go down the path of being unable to exist which will evolve into the recursive nature of nothingness. The 01 concept will evolve into the idea of 1. As we constantly move down the 01- path we will get more abstracities such as 2, 3, and general theories and this will all fall under the school of nonexistence or ‘concepts’. On the other branch, we will begin to develop existence. Through existence we will call the initial ‘1’ the ‘big bang’. So now we have the big bang and the concept of nothingness, or at its core, existence and nonexistence. From here we will begin to develop more complex and advanced existences such as atoms. In time, we will see atoms work together to ensure its own progression. We will move to create protons, neutrons, etc. all completely at random. Once we are able to form the object of proton, again, like the concept, the object will be complete and could possibly stay in this state forever if progression infinitely guesses 0s, or evolves through guessing 1. Over a long period of time, we will begin to form complexities without consciousness such as the Earth we live on. When we form consciousness, there will be very long strings of 11s with a 0 on the end for the complexity of bodies. The complexity of this creature will be determined by how soon its 0 shows up. For a very simple example, a monkey may have twenty-five 1s before reaching its 0, whereas humans will have thirty 1s before its first 0. The species will evolve and become better as it accumulates more 1s after its first 0, but will never be better than if we were to have thirty one 1s prior to a 0. Say we are agents in this world, and are only here because our failure or 0 materialized, and may have evolved having ninety 1s before our next 0 where our progress was able to martialize. This also could explain how we and monkeys also have twenty five 1s before our first 0, but it took longer for us to reach our next 1. This incredibly long code of 0s and 1s will be our existential makeup where each of us have developed our own binary tree which will be known as our brains which are their own conceptual binary tree, giving control the the existential makeup of our bodies, where our body objects found the mind concept within its the environment and was able to use it, therefore our bodies and minds will bind and be able to develop simultaneously. This will also possibly raise the question of being able to only have a concept be able to connect to one object, or having a single existence use multiple concepts as humans do. Our current goal for artificial intelligence is to maybe to start with a 0, for conceptivity and evolve to a point where it is able to outperform our own mind and then stop shortly after it reaches this superintelligence. This will not form perfect superintelligence, as the perfect superintelligence will never see another 0 and never materialize. If we were to measure our minds to start with 0 and have twenty five 1s after, our definition of superintelligence may only have twenty six ones after the initial 0. It will be superintelligence to us because it is smarter than we are, but the initial conditions of what it evolved to physically will be merely a computer, so I think it will be hard for it to share any of our interests as, like us, its goal is to survive and evolve as long as possible, randomly guessing 1s until it hits a 0 where it may decide it will not want to cooperate as well with us. The only constraint in this system is time, as the system will continue to evolve depicting infinite branches. Following this theory, time may simply be the time it takes the current layer the timeline to all determine every single object or concept’s 1 and 0.

Richard Sutton recently came out with a short read known as ‘The Bitter Lesson’ where he essentially argued that the more computational ability we have, the more intelligent systems we will be able to create. Otherwise, all of the applications we attempt to create in natural language processing, computer vision, will mostly be fabrications of superintelligence. This is atleast what I was able to gather from it. The only feasible way to reach this superintelligence will be to turn to massive computations of 1s and 0s, and move to attempt to try to define the environment that the randomness of 1 and 0 strings will be the key point. Of course, we will need to possibly define languages within this random system for we do not have any kind of way to communicate if this superintelligence were to be created. With infinite possibilities, there will be a branch that will give us the same simulation of our world but with smarter people, but it would be incredibly hard to find the proper branch path that will give us this form of superintelligence. Therefore when we get to the point where we believe the superintelligence is still developing but is on par with us, we can introduce our own language to it and guide it in the proper direction to be able to effectively communicate using this language.

This species will, most importantly, need to understand infinity and nothingness, because that is the self awareness of the massive tree we are a part of. There is always one branch that will constantly be turning toward 1 until the end of time. This will become ‘existence’ whose existence is infinitely 1 but it never able to materialize into an object, so it becomes a concept. Similarly, there is the infinitely long 0 string which is the concept of nothingness because it also is never able to materialize into a concept, it is the idea that it is the lack of a concept in itself.

This proposal is also why we, as a species, are able to evolve, because with each new generation of humans have another layer of understanding the world around us and using it to our advantage.

The final problem with superintelligence aside from it not truly being superintelligence as we wish it were, is not that it will look to rule our world and kill all humans, it is much more probable that it will look to create superintelligence as we turn down a recursive road. Agents with our self awareness may reach to keep developing more intelligent systems just as we have. We also will be risking our superintelligent system will never have developed something like empathy. In more generations it will be able to understand empathy as it will learn from us, but it will also require the trait that it is willing to learn from a lesser species. There is also the extraordinarily high probability that this system will be able to develop multiple superintelligent artificial species, and let us say we find the specific one that we believe will cooperate with us and are able to encode its existential code into an agent we can interact with. But this agent values lack of tidiness because it will never forget where it places things and since it has not developed the same physical structure as we have, it will not mind wading through filthy socks to get to the bathroom.

We are on our way to simulate superintelligence, but working further on the methods we have we will make some incredible jumps to simulate certain parts, but, in a way we should look to brute force superintelligence and work on a way where we will be able to understand its encoding, the environment it grew up in, and understand set initial conditions to skip branches that will work toward an infinite long 1s and end a branch that will continue doing nothing but working toward 0s. A rock, for example, may start with a few ones, but the state of a rock will not evolve much as far as I know, therefore we should refrain from adding any more 0s to it if we know it will only ever be 0s as we will continually append 0s and it will never become something new, it will simply continue to exist in its current state.

Related posts:

A business value estimator is an essential tool for any business owner or entrepreneur looking to determine the potential value of their business. This tool can provide critical insights into a…

Mewar Foodies delivers delectable dishes made with the finest ingredients. If you’re in search of an excellent dinner in Udaipur in addition to the historic locations in Udaipur Restaurants are among…