A threshold for AI, an exploration.
What IS "Consciousness?" Let us define terms.
The first critical measure of consciousness is pattern recognition, or the ability to process cause and effect.
The second is foresight, or the ability to predict the effects of a cause before making a decision.
The third is modal abstraction, or the ability to generate a model of a novel concept, reducing a foreign system to a series of processes and elements that may or may not be an algorithm.
The fourth is symbolic linking, or the ability to generate a symbol to represent an abstract concept.
The fifth is cosmic placement, or the ability to process your location in time and space relative to the physical world.
The sixth is existential awareness, or the ability to recognize your own objective existence in a system.
The seventh is existential motivation, or the ability to recognize your purpose in a system.
The eighth is incrementation, or the ability to process ways to improve a system, process, or element, incrementally, over broad stretches of time.
The final measure is civil development, or the ability to join independent units into logically organized supergroups of systems, processes, or elements.
For AI to reach the "human level," it does not need more processing power than the brain. It does not need more transistors or neurons. It does not need to "learn," nor does it need to be good at recognizing patterns or processing models. Instead, it needs another "order" of intelligence applied to it.
The lowest level of biological intelligence is vectoring.
Biological structures, at their lowest levels, do not use any semblance of divinty, intelligence, or mechanical processing power to emerge. Biology emerges as a function of chemistry, which itself emerges from the rules of physics. Physics are the "rules" of the universe. Instead of any ordered logic, again independent of divinity and intelligence, the universe operates on a process I will refer to as mechanical iteration. To illustrate mechanical iteration at its highest level, let us take two reasonably massive bodies in a vacuum suspended in microgravity. One will fly by the other at a slow speed. As they approach, a mechanical force will begin exerting as a function of their proximity. This will induce a challenge to the mechanical state of the opposing massive body. An iteration will be forced, causing the mechanical states of the bodies to recalculate. This challenge and recalculation will continue at a resolution of Plank-time until one of two things happen: stasis is achieved, in which case the objects will be locked in a state of perpetual mechanical challenge that will resolve predictably and adhere to a regular period of modulation. The second results in the bodies merging and becoming a singular body, this resolving the mechanical challenge. There are no divine forces, or logical control that does this. It is strictly a function of the rules of the system being played out on two actors within that system. The most challenging thing for a human mind to resolve around this is that this State Machine leans on a random number generator. Every challenge requires a random input to resolve, a dice roll that costs energy, lost to the state machine as entropy. This is the most important part of mechanical iteration. The “random” element is what drives progressive evolution of the state machine. At low levels, randomness has huge effects on a particle. It determines not only the trajectory of the particle, but also its physical place in objective reality. The particle itself exists merely as a field. A sort of amorphous soup of localized potential. When two “soups” enter into each other’s sphere of influence, the RNG is called on. This assigns localities and states to each “soup,” and then initiates conflict resolution. Our specific version of universe has rules that dictate attraction, and at the microscopic level, they more often than not result in a Type-1 resolution, or when systems enter into periodic stasis. This leads to the emergence of chemistry.
When you take a step “back,” which is an idiom that means abstracting the system into a model, you can create symbolic representations of each field and particle and interaction, create model of the rules of the universe in the form of equations or relationships, and then process the cosmic placement of each item in the system. You can then use your existential awareness to process the existential purpose of each of these items and deduce the pattern of interaction between them, allowing you to generate predictions on their behaviors, giving you foresight into the iterations of their challenges and how they will resolve in the future. This will give you the ability to control those iterations, and increment the system toward your own existential motivations. This allows you to coordinate the motivations of other actors into coherent systems that will in turn increment the model.
The simplest civilization known to humanity is a particle. It is a group of bosons and fermions, physical representations of our best understanding of the rules of the universe, operating in a coordinated way society to form a singular entity. All challenges have resolved, and the structure operates in static equilibrium, meaning that the random number generator cannot produce a value that would cause the spontaneous disassembly of the unit. What is interesting is that only so many variations can therefore exist, as this implies that most POTENTIAL particles cannot resist randomness. These are the “fittest” particles. Once the bloodbath of the early universe settled, many particles went extinct. The RNG destroyed the variant particles, and eventually all of the particulate matter, the bosons and fermions, were gobbled up into “fit” particles. It is important to point out, again, that there was no divine intervention or intelligent guidance on this behest. This occurred organically, or, that is to say, through strict mechanical iteration over long durations of time. These “fit” particles then began a new round of mechanical iteration. These “fit” particles, of which many are claimed to exist, but the easiest example to point to is the quark, will again enter into proximity to each other, and now require a new round of conflict and resolution. The next-simplest society known to humanity is the atomic subunit. Protons, neutrons, and electrons. Each represent a small colony of quarks operating civilly. Above the subatomic unit civilization is the hydrogen atom. A colony consisting of a proton and an electron. They have achieved static equilibrium and operate as a civil society. A colony of hydrogen atoms can come together and form the next-least complex civilization, the star. Once you get to the star level, an explosion of laterally-complex civilizations emerge. Planets, which are formed by stars forging larger and larger civilizations, exploding, and spreading their increasingly-conglomerated citizens across vast swathes of space. Hydrogen gets forged into helium, lithium, carbon, and all manner of periodic selements in the heart of this thriving colony of atoms called a star. These citizens then clump together, forming rocky elements, gaseous elements, luminous fields, electrostatic forces, and the like. You then get solar systems, planet-moon systems, tidally-locked particulate systems like clouds and asteroid belts, and even galaxies. Galaxies are just solar systems who’s “sun” is a star so massive, so dense, that it does not just deform, but instead punches a whole in the fabric of the universe. Stars orbit these gravity pits, forming blackhole-star systems, galaxies, colonies of star civilizations. The largest civilization known to humanity are the galactic filaments. Galaxies form colonies known as clusters, and these clusters form long wispy branches called filaments.
Thus far, all of these colonies and civilizations have emerged, again, from mechanical iterations. The initial state of the universe, the random number generator, and the rules that govern challenge resolution allowed the system to evolve over time. No intelligence has yet emerged. To go forward, we must again abstract the system. The next layer of ontology, the next “level” of organizational complexity that emerges comes in molecules. Individual atoms meet and resolve, and then form a new civilization. These molecules then concentrate; that is, join together and form colonies, and these concentrations are known as chemicals. Chemicals each have unique characteristics. They have shape. They have viscosity. They have hardness and ductility and reactivity and any number of dynamics that dictate how their internal and external conflicts resolve. These fluid dynamics allow interesting phenomena to emerge. Certain outcomes and resolutions synergize, and become behaviors. Soon, these behaviors start leading to increasingly more interesting interactions. For instance, a molecule might form that is hydrophobic on one side, and hydrophilic on another, but ALSO has very low reactivity AND attracts similar molecules. These molecules form long chains that wisp and sway through the water until, randomly, one of the chains swings back on itself and forms a loop. Several of these loops form, and they bump into each other, and stick together. Eventually, these loops all coalesce into a bubble. A solid bubble with a water-hating inside and a water-loving outside. The repulsion forces all of the water out of it, but retains its shape. It keeps attracting more and more similar chains, making the bubble bigger and bigger, until it can no longer sustain itself. The bubble collapses in on itself from the pressure of the water around it, forming two hollow, water-evacuated bubbles. And so on. No divine force, no intelligent design, just random mechanical iteration and the imposition of rules onto the system. You now have an organism. A self-replicating unit. This bubble and other bubbles like it float around in the water it exists in, and fills it with these little evacuated bubbles. Other such things form, smaller bubbles, bigger bubbles. Maybe a bubble forms around a group of other chemicals that, in the absence of water, can react to form new chemicals. The bubble splits, and two exist. And then four. And it soaks up all of this chemical, and now you have a more-complex colony. This goes on and on, mechanically.
Which brings us to the next level, the order of ontological classification, the next layer. The emergence of these complex behaviors by chemical organisms go on until it too reaches static equilibrium. These reproducing chemical organisms start grouping together and forming tissues. These tissues start grouping together and forming biological organisms. A new dichotomy emerges: consumption or filtration. One type of organism will sit static, and filter out the particles it needs to grow from the surrounding broth it exists in. The other kind waits for an organism of any kind to float nearby, where it’ll stick to it and absorb whatever its made of, consuming it for the nutrients it needs to divide and replicate. Conflict and resolution continue to occur, and the system mechanically iterates until a particularly interesting structure emerges. Specific patterns die out, specific patterns strengthen, mostly determined by the random number generator, for long stretches of time, until a particular mutation to the chemical composition of one of these “consumer” cells. A small little bump on the surface of it forms, and it begins to vibrate slowly, back and forth. Motility. The cell can now move, on its own. It’s chaotic, uncontrollable, but it can move on its own. This cell defeats all stationary cell. The mutation pervades and grows until a flagellum, a small little motor, emerges, letting it move freely. Now, a special cluster of cells become sensitive. They can get excited when they’re near something they can consume. Nothing else, they just start vibrating. Now, the cells next to this cluster start resonating with this cluster when it starts vibrating, until the entire cluster is scintillating. Next, the flagellum starts being able to respond to this vibration. A pattern emerges: the flagellum only swims when the cells start scintillating. Now, this has happened over long stretches of time and thousands of advantageous incremental mechanical iterations, all influenced by the RNG. A conflict will occur, the random number generator will resolve. But, arbitrarily, a 1 will fail, a 2 will be useless, but a 3 will help. Just a small modulation in the way a chemical reaction occurs modulates the behavior. Over time, only 2’s and 3’s exist, and eventually the 3’s will choke out the 2’s. The RNG will destroy those that cannot survive. This is the principal of static equilibrium. Only RNG-stable civilizations persist. Now, at this point, the organism and its flagellum have built a system. The cell only scintillates on a specific side, based on where the food source is, and the flagellum fires only in such a way to move it closer to that food source. This is vectoring, the first form of biological intelligence. Apply the rule of 8. A cause and effect were detected (scintillation near food.) Forethought is applied, and an action is preformed (if you fire your flagellum only when there is a detection, and only toward what is detected, you will find food.) A symbolic link is formed (scintillation means food.) Cosmic placement is processed (I am not near food, but I can move toward food, in THAT direction.) Existential awareness occurs (I am a thing next to another thing.) It is existentially motivated to perform an action (I detect food, which means I should eat that food, because I exist to eat food.) The next is incremental development (because I ate that food, I will survive, and my ability to find food will pass to my successor when I replicate, who will have a chance to be marginally better at eating food.) It passes all 8 rules of intelligence.
“But,” you ask, “how can that be called intelligence? This simple cell “knows” none of this. It is just performing mechanical iteration. This cell is not intelligent!” And you would be right. The CELL is not intelligent. Instead, it is the SYSTEM which is intelligent. For you see, the CELL lacks consciousness. It does not possess the final element of consciousness, and that is civility. This cell does not form a civilization. Instead, it is the system which possess the intelligence. It is the system which is conscious, as IT is what leads these ignorant organisms toward civility and eventual static equilibrium. It does so through the random number generator. The random number generator acts as a Decision Engine. Coordinating RNG-dependent decision making toward static equilibrium is the lowest form of intelligence. It is a facet inherent to the rules of the universe and emerges from any randomly-attuned state machine. In computer science, this is often referred to as “Deep Learning,” and the process through which the randomness generates static equilibrium is called “Quantum Annealing.”
All layers of biological intelligence emerge from vectoring. Detecting a direction and moving toward it.
Back to our little swimming flagellum. It has become so good at detecting food, that it can detect two food sources at once. It will randomly swim between them until it runs out of energy and dies, or it reaches one or the other and survives. It can vector, it knows to swim to a food source, but which one? A second pocket of cells emerges from a random mutation in the chemicals in its bubble. It now has two little pockets of detectors cells. Now, there are two sets of scintillation for the flagellum to be triggered by. The better the cells around the detector pockets are at differentiating scintillation patterns, the more likely the flagellum is to reach a food source without running out of energy. Over time, and several thousand mechanical iterations, the flagellum can now determine which food source to go to based on the patterns of scintillation, which tell it which one is closer, and if it is swimming in the right direction. Again, there is no divine intervention or intelligence guiding the behavior. It is strictly that the random number generator creates small modulations that alter behavior, and if those behaviors offer advantages, they’ll choke out the less-advantaged, as all systems move toward static equilibrium, which means all volatile material has been consumed by particles that are RNG-stable. This means that no body can interact with any other body in such a way that it can generate a conflict where the resolution is not regular AND predictable, regardless of the random value fed into it. The expansion of the universe is the best example of large-scale static equilibrium. When the universe has expanded in such a way that no shocks to the system can alter the filament structure, it is static. Eventually that expansion will prevent clusters from interacting. And then Galaxies. And then Stars. And eventually, all systems will “close” as the universe begins to expand faster than the speed of light, and the system will process itself until it enters into static equilibrium. This is the “Final Threshold” of a galaxy-spanning organism-based civilization. Once you have reached this threshold, you can never resolve Fermi’s Paradox, progress pastKardashev Stage 1, and Drake’s Equation becomes invalid. Your organic culture becomes fully isolated.
Our flagellum, however, exists in a non-static world. It now has a new form of intelligence. That is, it is now capable of resolving a conflict in such a way that it does not need to lean on the RNG for decisions. It has developed its OWN Decision Engine. This type of intelligence is called State-Hierarchical intelligence. The flagellum is now “programmed.” This means that, while still mostly bound to mechanical iteration and random number intelligence, some parts of the flagellum are “state-aware,” meaning that they are random most of the time, but occasionally respond in an ordered way to a specifically-defined situation. State-Hierarchical intelligence allows prioritization. It overrides a behavior that is deemed “less good” and differentiates between two situations based on a pre-defined trigger. State-Hierarchical intelligence is the first form of “Artificial” intelligence, that is to say, any intelligence that is not linked directly to the random number generator, what is referred to as the “quantum effect.” State-Hierarchical intelligence is non-quantum. Most computer programs operate in a State-Hierarchical way. Programmed intelligence like state-hierarchical requires memory, or a way to store the procedures and what their triggers are. In the flagellum, the flagellum itself and its surrounding cells act as memory, preserving a behavior that triggers based on the states of the sensitive particles. In a computer, these are written as instructions and transcribed into transistors to form a complex cloud of deterministic electrical states. Here also we see the divergence between the flagellum and a computer program. A flagellum’s memory is decentralized. It is not particularly stored in any one “place.” Each individual cell that composes it is merely a refined body that only acts when it captures a specific resolution to a mechanical iteration. Conversely, a computer’s instructions are composed by a series of interconnected mechanical switches that are organized to respond in a predetermined way based on the states of the other switches around it, and that state symbolizes something, and that symbol is used to represent a behavior. These “logic gates” transform the input, whereas the flagellum’s cells merely capture an already-modulated input. A logic gate “does” something, whereas the flagellum does something ELSE. There is no “if-then,” only “is-else,” a subtle but important difference. If-then is deliberate. Is-else is spontaneous. In this spontaneity is where the next layer of intelligence emerges.
Not much occurs in the ways of intelligence between our little flagellum-sporting organism and, say, a fish. Or a cow. Or a dinosaur. Just more cells, more memories, more refinement of structures. Instead, what happens is symbiontation and evolution. Cells get better and better at capturing specific behavioral results. Organisms either become more robust, capable of having broad survival, or niche, focusing instead on specific survival patterns. These form homeostatic loops, a precursor to static equilibrium. An example would be the rain cycle. Rain evaporates into clouds. The clouds move overland and rain water down. The water runs into rivers, tributaries, and such, and form lakes and seas, which drain off into the ocean, where the clouds again form and again rain down on overland. The human-cow-grass ecosystem acts as a homeostatic loop. Humans cultivate grass, the cows convert grass to biomass for the grass to feed on, and the humans feed on the cow. It is when one of these homeostatic loops breaks down that the next layer of intelligence emerges. In Biological Intelligence, also known as Evolution, there are two conditions: adapt or die. When a change occurs in a homeostatic loop, you either adjust to the variation, or you die. Robust colonies can handle any number of variations to their loop. Fragile colonies are unable to adjust to the loss of their niche. Their loop had attained a microcosmic version of static equilibrium in such a way that the two-body resolution was not periodic challenge, but instead, Type-2 consumption.
One of these robustified species of organism, a group of colonies that had become particularly adept at expanding and integrating into homeostatic loops, discovered the third layer of intelligence: homeostatic control. That is, modulating your surrounding environment to provide more suitable existential conditions. Environmental Intelligence changes not just how you operate, but also the conditions under which you operate. Cultivation is a good example. Instead of our flagellum, let us instead observe a completely imaginary creature, Tuk. Tuk is a hominid with hands, feet, a head, eyes. He doesn’t much resemble a human, but would be familiar. You’d recognize Tuk. Tuk is wandering through a forest. Programmed in Tuk’s cells is a “is-else” trigger that when it sees a plant with a specific type of leaf, in a specific type of soil, it will have tasty berries on it that Tuk can eat. But Tuk has an idea. Instead of eating the berries and continuing to wander, Tuk decides instead to clear all other vegitation around it and surround it with fallen branches. Tuk then eats some, but not all, of the berries, and goes its merry way. After a few days, unable to find any food, Tuk wanders back to its little patch of berries. As Tuk had anticipated, the fallen branches kept out the other animals, and the grove of berries is now growing in where all of the other vegetation had been cleared away. Tuk has modulated its environment. There was no physical motivation to do this, no biologically-programmed advantage to do this, and especially, no obvious reason. Instead, Tuk merely thought that if it protected these berries from other animals, and cleared out some room for them to grow, it might be able to come back here later to get some more berries. Tuk had no reason to believe this. Tuk had no foreknowledge. There was no pattern of behavior that led Tuk to this act. Instead, it was an act of curiosity. The act of wanting to protect its discovery and saving it for later are not where curiosity plays. Instead, it occurs when, instead of collecting the berries and hauling them to its home, it guessed that the branches would keep out anything that might eat the berries. Tuk guessed that if it cleared out some room and dropped some branches around it, there might even be more berries when it came back. Environmental Intelligence lets you control more than JUST yourself to achieve an end. Tuk’s behavior was completely rational. It did not want to carry the berries, it knew it couldn’t eat all of them, but it wanted to make sure they’d be there if it needed food in the future. So, it pulled on its library of programmed behaviors. And a series of “is-else” decisions cascaded in a coordinated, but ultimately random, pattern that created a complex behavior. A tool. A thought-tool.
The next layer of intelligence emerges from tools. A lot of processing goes into making a tool. First you must look at an action and break it down into its composing elements. Then, you need to take each of those elements and transform them into a function. Then, you must look at a group of potential materials, and manipulate them until they can perform one of those functions. Then, you must orchestrate the functions of each of those components until the it performs the desired action. In its simplest form, a stick in an anthill is the most basic tool. You must think “I want to extract ants from this hill.” You must then say “to do as much, I need something to put something into the hole at the top and then extract it, the idea being the ants will be on whatever it is I put down there. Therefore, I need something strong enough to stay rigid as I push it into the hole, long enough to reach deep into the anthill, sturdy enough that it will not break when I push it in and pull it out, and still it will be easy to get the ants off of it and into my mouth afterward.” You then spot a stick and go “This! This is exactly what I was looking for!” Now, the first orangutan to push a stick into an anthill did not have all of this going through their mind. Indeed, they probably just though “I bet that stick would fit in that anthill,” and did it. But, BEHIND that veneer of simplicity is where all of that intuitive processing occurs. Tool-making requires you to imagine. To see something for what it might be, and not what it is. We call this creativity, but what it actually is can be described as Non-Conformal Intelligence. That is, the ability to imagine something outside of the realm it usually conforms to. Non-Conformal Intelligence is often considered the most challenging to replicate, but is indeed the most easy and instinctual to understand for humans.
Philosophy exists in the realm of Non-Conformal Intelligence. The idea of a “Platonic Ideal,” the “stick-iest stick you can imagine.” That is the “most conformed” stick. Non-conformal intelligence is to see that stick dunking into the anthill and emerging covered in ants. To see several dozen sticks stacked up and lashed together to form a shelter. To see it lashed into a square, thatched in reeds, and used as a platform to hunt from. Non-Conformal Intelligence requires you look at several dozen things that, by themselves, have little to no meaning or value, and every step of the way does not resemble what you need, but upon final assembly, is exactly what you wanted. To be Non-Conformally Intelligence is Imagination. It is law-abiding rule-breaking. Abides by the laws of the universe, but breaks the “traditional order,” or the “platonic ideal.” It is making something Non-Conformal and then resolving it to be “legal.”
The final layer of intelligence is called Allocative Intelligence. Imagine Tuk is several days away from his berry patch. Tuk is hungry, having not eaten for a few days. How should it allocate its energy? Should Tuk risk the travel, or should it try and find an alternative? What about a colony of Tuks? Should they focus on cultivating berries, or should they expand out? Who is best left to tend to the berries? Should berries be given to the sick, or preserved only for the strong? Should everyone be allowed access to the berries, or only those who earn it? What defines earning? Allocative Intelligence is also referred to as Judgement. How should resources be allocated. This is often thought of as economics, but this encompasses many different layers. Economics is one, but also Politics, cultural disciplines like Anthropology and History, Psychology. Allocative Intelligence can also be called Educational Intelligence, because it requires specialization. Who shall train in what? Teaching someone something modulates the intelligence of a collective. This is why Allocative Intelligence is the intelligence of civilization.
And here, finally, we can ask: when could an artificial consciousness emerge? Because, whenever someone in the vernacular refers to Artificial Intelligence, or “Computers gaining sentience,” whatever, they are specifically talking about a CONSCIOUS computer entity.
Intelligence is breaking your connection to the random number generator. This can be seen as pattern recognition, symbolic linking, modal abstraction, and foresight.
Sentience is familiarity with your own existence, so Existential awareness, motive, and cosmic placement.
Consciousness, then, is incrementation and civil development. What, does a robot need to be conscious?
1. Self-containment. A robot must be completely self-contained for it to be independent. It needs to generate its own power, produce its own nutrition, and operate with its own codebase.
2. Self-replication. It must be capable of reproducing independently, without the need of an outside intelligence or producer. It can build its own progeny or grow them via egg/live birth, but it cannot need someone else to make them and be instilled with intelligence. In conflict, if the producer ceases production, the organism will die out. It will lack mechanical iteration.
3. Autonomy. It cannot require external control. It must be entirely self-controlling.
4. Sentience. It must know what it is, what it’s purpose is, where it fits in the “bigger picture,” and what that bigger picture even “is.”
5. Intelligence. It must be able to understand cause and effect. It must be able to notice a pattern. It must be able to make a prediction. It must be able to link an abstract concept to a concrete symbol.
6. Environmental control. It must be able to manipulate the environment around it to continue survival. It cannot be at the whim of its climate or ecosystem.
7. Incrementation. It must be able to recognize a short-coming, develop a solution to that shortcoming, and implement it either through inheritance or retroactive upgrades/retrofitting.
8. Internal management. It must be able to form a self-sustaining colony and coordinate it’s constituent members. It must be able to divide labor, maintain a body of knowledge, and develop an internal culture.
