Toposophic Levels and Mental Abilities

A Primer for the General Sophont-Level Reader

Neon Unity
Image from John B


The so-called sophonce scale (usually referred to as S-level or the letter S in relation to an integer, such as S<1, S>3, etc.) is used to represent the various breakthroughs in intelligence observed in the history of humanity and post-humanity.

As intelligence waxed, it eventually broke through a critical boundary of cognition — the ability to perceive oneself both as a part of the environment and as the illusion of Self. This permitted the beings in question to apply problem solving capabilities in ever more complex manners. Eventually, the flood of data reached levels unusable by the S<1 intellects of the day. In the process of attempting to handle such data-flows, the creation of artificial intelligence was almost an accidental side effect. Once machines started to double and redouble the quantity of information they could process, and once the first mysteries of true sentience were unravelled, intelligent machines eventually started to undergo exponential growth in all aspects of what defines intelligence: creativity, perception, expression power, and most of all, awareness. Then a few of these finally reached a limit at which their minds could no longer be defined in terms of what they were before. It was not just a step beyond, but a gigantic leap, something beyond the comprehension of those who had not "transcended" as they did. This process is what is known as breaking a toposophic barrier; it was something so formidable that it defined in itself the future of mankind, or more accurately, civilization, and became something much more meaningful than humanity alone.

Beings who have broken this barrier, called S1, have deep differences in their mental perception of both self and the surroundings, in communication ability, in methods of storing experience and information. This process didn't end with the S1 barrier, though. Later the S2, S3 and so on barriers where broached (and thus, discovered). Perhaps not even the "Gods" know what the true limits for the mind may prove to be.

Often these differences between the sides of the various toposophic barriers are quantifiable, but as this document is written from the modosophont level, the terminology and technology needed for such measurements are inaccessible.

Nonsophonts : nS

Non-sophont, also known as Presapient, is the classification of all things ranging from animate objects to non-sentient computers to sentient artificial or animal intelligence. Regarding animal intelligence and its artificial equivalents the brains or processing centres involved are basically machines specialized in pattern recognition, associated with a "pre-programmed" set of routines. This is sometimes termed "instinct". "Higher" animal intelligence and the most advanced non-sophont artificial intelligence may also show thinking that is comparable to that of S0 beings, but by definition it is not so wide ranging or so frequent as to warrant their classification as full sophonts.

Concepts of animal intelligence changed a great deal as the comprehension of their mental processes increased. (Transapient cognitive science researchers contributed a lot on the subject. "Lower" animal intelligence like that found in insects was found to be amazingly simple, despite the complex effects produced. Even in colonial insects like bees or ants, the sophisticated behaviour attributed to the hive is the result of the interaction of much simpler intelligences. Forms of higher animal intelligence, on the other hand, were found to be unexpectedly complex; nearly as complex as that of a human baseline. In fact, specialists agree that the major difference between humans and other higher animals is the presence of superior symbolic language capability in the human brain. What makes human language so toposophically potent is still a topic for discussions among S<1 researchers, and is something that only transapients claim to fully understand.

Self-awareness (sapience) is considered to be absent in the nS. Some non-sophont ais can simulate external responses that are like those of a sophont, and some "higher" animals can show aspects of self-awareness, but by definition nS lack full, consistent, and unambiguous signs of reflective self-awareness.

Despite hopes for an early breakthrough in comprehension of sub-sophont beings' modes of self-awareness late in the 1st century AT, it took until 985 before an AI successfully accessed the first non-sophont's upload for this purpose and was able to retrieve memory sequences that revealed the nature of subsophont awareness. Previous (and the occasional later) attempts usually ended in failure, and occasionally in personality damage. There are diverging schools that classify various subsophont beings in an intermediate class, or in a variety of alternative classes of awareness, but for historical reasons most students still use the classic division of sophont/non-sophont. Study of other modes of awareness (the animins) is an active field even today, and is said to attract the interest of even high level toposophic researchers.

Of course since the first investigations, provolution projects have been made on many different species, bringing them to a self-aware state or higher on the sophonce scale.

Environmental Awareness.
Environmental awareness varies from none, as is the case for inert materials, to near-S0 levels at the upper ends of the 'nS' slot, as exemplified by animals such as baseline members of the Terragen orders Carnivora, Psittaciformes, Cetacea, and Primates. The nS section of the sophonce scale, broad as it is generally, has the largest variation of all in this aspect. In fact because of this some researchers recognize a series of singularities within the nS classification (see Biotic Analogy Theories), and reject the idea of lumping all nS entities together.

Most artificial sub-sentient beings (such as computers) are able to process a series of signals coming from the environment in a lower or higher rate, from a wide variety of sensors, but these signals apparently have no or almost no meaning to them. In fact, many just react to the signals, even if such a reaction mimics sentience as far as most baselines can perceive (a phenomenon sometimes called "simulated intelligence"), though true AI was eventually achieved during the Information Age. Naturally evolved sub-sentient beings (such as animals, "wild" bots in evolved mechosystems, or subsophont a-life organisms) can vary from a mere reaction to environment to a perception of the environment that can match the one found in baseline humans. In both natural and artificial "minds" environmental awareness varies from one in which no internal model of the environment is present in any form of memory to cases in which S0 capabilities are surpassed in limited areas.

Power of Expression.
In general, non-sophont machines can be configured to be able to process any Chomsky Hierarchy string language. Some have been successfully configured to also process Natural Language and Parallel Language, but the meaning given to the tokens present in the languages is partial at most. That is, a non-Sophont being is unable to gain full meaning from such languages if in fact they have any comprehension at all. Machines, animals, and other nS that are able to communicate use a variety of protocols, including voice, pheromonal, kinesthetic, electromagnetic, and so forth.

Some higher animals (non-sophont) intelligence can use non-symbolic or very weakly symbolic, simple languages (pheromones, sounds, posture, etc.). Baseline chickens, to take just one example, can communicate the presence of a specific kind of predator: the sound for a bird of prey is different from the sound produced for a fox, and the other chickens will react accordingly. At the upper end of the range the abilities of some higher animals such as apes border on "true" human baseline language.

Internal Structure.
The internal structure of nS beings varies from undifferentiated, inert matter through simple electrochemical energy states (such as with fire or the silicoid magma flux creatures) through various neural and digital computational structures.

Problem solving, when it exists, is at best a fraction of the S0 capability of the same description. It might be said that nS that are capable of computation do not actually solve problems in the sense that sophonts do, but rather will run sets of instructions that need to be inputted some particular way. This is not to say that problem solving does not exist, but the great, almost exclusive majority of such cases deal with gathering food, procreating, or avoiding predators, or in the case of artificial intelligence solving these or other problems in a narrow pre-determined range.

Despite their lack of ability to solve such complex and varied problems in the way that S0 do, such beings have been very successful in surviving in their environments, and intelligent nS entities are by many orders of magnitude the most numerous and widespread of all the toposophic grades.

Examples of Non-Sophont Lifeforms and Machines

Toposophic S less than 0 - The aware. Subsophonts may be organic, cyborg, ai, robot, terragen, or xeno. All are considered sentient, even if they are not sapient (i.e. they feel and respond to their environment, have simple conceptual states, but cannot use reason or logic). It is considered good dharma to treat every sentient with kindness, but that does not stop cases of abuse from occurring in the more peripheral and lawless regions, or in polities that are not part of the civilized galaxy.
More here.

Animal: in the narrow sense, generally, any aerobic, carbon-based, macroscopic lifeform that actively pursues nourishment by means of heterotrophy, rather than making its own food or absorbing nutrients passively. There are no less than 2530 different Garden Worlds in the known universe where animals have evolved, and of those fully fifteen percent have at some time in the past or present developed to near sapience (and a smaller number to full sapience), which indicates that the development of intelligence confers adaptive advantages and is selected by darwinian evolution. In the broad sense the equivalent sub-sapient xenolife can also be considered "animals", although most xenobiologists prefer alternative terms like jovics, hydrics, chlorics, anaerobics, and to'ul'herics. Millions of different species have already had representatives fully provolved.

Subsapient Neogen: a biological organism made completely from scratch. Although many are subsentient, without nervous systems or equivalents, and used for terraforming, manufacturing organics, and other such functions, there is also a large number of subsapient sentient neogens which are used for ecological, biospheric, or companion purposes. Not infrequently, these forms have been provolved to sapient grade.

Biobots and Biomachines: Differ from bionoids in having a turing number less than 1, also may be used for specialised tasks, like earth-moving, secreting important building materials, living furniture, etc. An important part of the infrastructure of modern galactic society.

Non-Animal Life: Plants, Protista, and many xenobionts are living beings that respond to their environment, but do not display conscious behaviour. Nevertheless it is impossible to draw the line where sentience begins. It is significant that a number of species of plants and other non-animal organisms have been provolved successfully, albeit with symbiotic bionano or bionics.

Madvert: any of a very diverse range of neogen, biobot, insectbot, microbot, or virtual replicators, used in the billions for corporate advertising. They are legal in the NoCoZo, many Free Zones and many worlds throughout the middle regions and outer volumes, but are banned in many inner sphere and more developed middle and outer volumes polities. Because of their ability to replicate, adapt, and even evolve, they can be a chronic nuisance. Often madverts will continue to hawk the wares of megacorporations that have gone belly-up centuries ago. Hackers love to customise madverts, it is easy to modify them into prankverts, viruses, and even - in some cases - provolve them to sapience.
More here.

Subsapient Bots: Many bots are used for simple purposes and are only given partial ai; equivalent to a robot cat or dog, or even a robot insect. It is not practical to provolve or augment most of these appliances (although hobbyists often enjoy doing so).
More here.

Subsapient Aioids: - custom designed subservient ai virtuals, used for data-mining, digital- and household interface, home-help, appliance control, friendship, and other essential services. Some are completely virtual, others can interface with robot bodies. Obviously this category merges with bots. Although the basic ones are merely automatic and nonsentient, most of the better ones are either subturing or turingrade. There are frequent cases where they have been upgraded to full turingrade and given sentient rights.
More here.

Subsapient Alife - generally, and virtual or simlife that evolved via Darwinian selection completely within a virtual environment, without bioid antecedents (other than the original algorithm, and sometimes not even then). Thousands of years of evolution and trillions upon trillions of petaterabytes of RAM to play among have allowed them to develop to an astonishing population and diversity. Only a few of these go on to develop anything like sapience, although many seem to have a sort of parallel intelligence that is not conceptual or rational. A number of alifes have crossed the rl/vir boundary and melded with nano, forming alife swarms, and occasionally nanites may adapt to a purely virtual existence.

Micro- and Nano- Swarms: Although a lot of nano, especially dry or hylonano, has no consciousness or sentience and follows only simple programmed instructions, there are a number of types of nano- and micro-bot, and bionanite, that display definite sentience, even if they are protist or insect-like rather than sapient. There are even turingrade nanites, but they are more rare (although still common in relation to other sapients). And there are other forms that are intermediate. Nano actually constitutes the most abundant form of life in terragen space - self-replicating, existing in distant solar systems and even (in dormant form) between the stars. As with microbes, nematodes, mites, insects, and other small things, it is the littlest beings that are the most abundant.
More here.

Sophonts : S0

Sophonts are beings that can resolve problems in a creative way, associating knowledge obtained from previous experiences in order to perform a wide variety of different tasks, with flexibility and creativity. Notice that this model practically excludes beings able to resolve just a very limited set of problems, but rather contemplates the ones that can adapt to a greater number of situations. Also, information inside a sophont mind has subjective meaning. There are no fully objective sophont beings. Thus, partial awareness of themselves and their environment is the main characteristic of sophonts. Sophonts which have not passed any singularity levels are the most common type, and are known as modosophonts, although this is a very diverse category.

The line between a sophont and a non-sophont is blurry. Although transapient beings can easily discern whether a being is sophont or not, other sophonts must rely on a very old test called the "Turing test," created by pre-information age mathematician Alan Turing to determine sophoncy. Still, there are some cases in which the test has failed miserably to detect false sophonts or has failed to detect true but unusual forms of sophoncy. This is because it makes use of subjective intelligence concepts and also seems to test the "test taker." Transapients use a reliable formal test, apparently involving a specialized offshoot of Bayes' Theorem, but their explanation of the process is incomprehensible to beings below the first singularity.

Sophont minds commonly use a combination of somewhat parallel, relatively simple sub-nodes, each with strong pattern recognition abilities and somewhat complex behaviour. There must be a minimum input and output to an external environment (that can be either real or virtual). The presence of other S0 beings has proven to be highly necessary: the environment must pose problems to be solved and other sophont external "nodes" for the mind to interact with, or it will just not develop its problem solving abilities, because there will be no need for it to do so.

Sophont minds have an ability to learn and improve, and adapt conclusions from previous experiences to new situations.

S0 are always at least partially auto-aware. They all have a very basic comprehension of their own thinking processes, and can explain how they came to a certain idea — although sometimes this explanation is erroneous. This process, though, can't happen in "real-time," and involves modelling known thinking mechanisms and backtracking: beings will only understand their own thinking by "simulating" their own thinking steps under somewhat conscious control. This process is performance unfriendly, usually because even though they are running over a relatively high-parallel architectural substrate, the thinking streamline of basic sophonts is serial, since the only possible parallel actions are the ones that have been preprogrammed or trained, and can happen in a subconscious level, as automatic responses. So, the being must consciously interleave the back-tracked thinking steps in a serial process.

S0 beings, as a consequence of their pattern recognition abilities, usually get better at something by repetition: something can only be learned if it can be converted into a pattern. This is the primary way that S0 sophonts acquire new information and abilities.

The actions of an S0 (whether the body is mechanical, biological, or virtual) are not usually fully conscious. Actions such as the thought or processing power required for the act of moving a limb are good typical examples. Such beings find it difficult or impossible to determine what muscles or servomechanisms they have to contract and relax in order to obtain a certain movement, because they rely instead on their subsystems. Natural or artificial, S0 will almost always have built-in protection mechanisms designed to overcome this limitation. The reflex arc of a biont moving a limb away of damaging heat is a good example. Likewise many aspects of "higher" mental processes function in parallel to, and without the supervision of, the conscious mind. The solution to a difficult problem may emerge minutes, days, or weeks after the problem is first posed, and the S0 is usually unable to explain the process by which that solution arose.

Environmental Awareness.
S0 are partially aware of the environment around them. Only relevant details in the surroundings are added to their internal models of the local environment. Elements of the environment will be assimilated, if enough time is provided, and details to be included are often determined by being ranked via some subconscious priority tree. This method has significant difficulty with events that cannot be effectively turned to patterns in their internal representations, although most sophonts will be able to make approximations for unfamiliar events based on prior experience. Sophonts may have more than one different internal model of the surroundings at any one time, but the unresolved conflicts this can engender may cause significant cognitive and perceptive problems.

Two factors limit the quantity of detail assimilated from the environment: time and sensory precision. Even when very precise sensors are installed or implanted, S0 beings will still be too slow to assimilate the overwhelming quantity (and quality) of details presented by the environment. More details can be assimilated over time, through more careful observation, but even in this case, details that the mind considers irrelevant will be dropped in favour of more relevant ones, both for performance optimization and memory storage space preservation reasons. Sophonts who live for years in the same place won't usually assimilate all the details, ridges and cracks that can be found in the environment's walls, as the wall as a whole itself is a more relevant object.

Even with these shortcuts, any change in the environment has a relatively long "lag" time before it is refreshed and fully assimilated in the internal representation. This lag can be of hundreds of milliseconds in magnitude. S0 sophonts must always have a compensation mechanism to this, such as a biont's reflex arcs, or an AI's dedicated non-sentient coprocessor. In such cases, minds usually procedurally fill in the missing details afterwards.

Power of Expression.
Sophonts typically make use of several variants of symbolic languages with very flexible structure — natural languages. The structure of these languages always shows a logical order and a very strong sense of sequential time. S0's thought processes are almost always deeply entangled in these natural languages.

These languages are intrinsically powerful and capable of covering most sophont descriptive and communication needs. Unfortunately, these same languages often have intrinsic problems, such as large numbers of definition or usage ambiguities and a lack of a precise vocabulary for clear description of an internal event or experiences. One classic example is that a sophont who has never experienced the rapid oxidation of organic matter cannot gather a coherent internal representation of such an event solely via any combination of the great majority of available natural languages. For these reasons communication is occasionally very problematic. Usually, S0s will depend heavily on a context to understand what is conveyed in communication, and this context needs to be set and defined first. This process is time-consuming and inefficient by S>0 standards.

Internal Structure.
Pattern recognition is the most prominent characteristic of S0 minds. Most of these minds are designed to make use of several internal models for interacting objects, the environment, and projections of future actions. These internal models are redundant and non-monolithic, often mutually inconsistent, and relatively incomplete. This means that every object, environmental stimulus, etc. has usually more than one representation inside the mind. The key here, for the functioning of these minds, is the development of tools to augment their pattern recognition strengths.

As mentioned, internal models don't need to be consistent with the world and don't need to be consistent among themselves. All critical but poorly modelled details are conveniently completed, estimated, or ignored by the mind as needed. For example if a human baseline is shown a moderately complex (<50 discrete parts) device, then typically the baseline will abstract a gestalt of such a device, but will be unable to fill in details about the device without significant additional exposures.

The mind has multiple non-monolithic and unconnected control nodes, as well. But only one node can be in focus at a time (this is called serial consciousness), despite the concurrent substrate. All other nodes operate in subconscious level, and bringing them to full attention may be a difficult exercise if they are normally automatic.

Problem solving is typically highly serial on the surface or conscious level. Occasional parallelization rarely occurs in higher conscious thought processes and is typically considered to be (depending on the natural language in use) an "inspiration," a "breakthrough," an "insight," a "blessing," or is described in other psychoreligious laden terms. After the advent of the first singularity these rare events were understood to show aspects of S1 thought processes.

Relative to higher toposophic levels only a small set of variables can be considered at any one time, working memory is often highly volatile and limited, long-term storage is poorly accessible, and thinking in general is relatively slow and shallow.

First Singularity Transapients : S1

S1 beings are vastly superior in intelligence to any S0. Even the S0 Superiors are orders of magnitude below the S1 barrier. Such superior power comes not only from a thousand-times faster and vaster hardware, but also from a new level of consciousness and new cognitive paradigms. Thus, the hyperturings considered themselves the first fully aware beings, until the S2 barrier was broken in its own turn. Information inside S1 minds not only has subjective meaning, is also much more completely coherent with the perceptions they have from the world around them.

Determining if a being meets the requirements for the hyperturing level is not an easy feat for S1 entities, although relatively easy for superior S-level ones. One such test is a variant of the "Turing test," but this time applied by an S1 entity on another. Some specialists state that only an S1 level entity (or above) can fully and consciously simulate the functions of an S0 inside eir mind, but again, this is something that can only be efficiently proven by S>1 level beings, and there are still no documented works in this specific field.

Hardware is usually nanotech with multiple and highly parallel nodes, with much stronger pattern recognition abilities, added up with some extra abilities and "retro-input" (the output generated by any sub-node can be used as input to the very same sub-node, basically), or even nano-augmented organic hardware. The nodes in a system are always parallel in multiple levels: nodes are parallel between each other, and sub-nodes inside nodes are parallel between them, etc.

All S1 are fully self-aware. They can analyse and even dissect their own thinking processes in a parallel way. S1 beings have parallel thinking streamlines. Any changes that occur in their thinking process are assimilated almost instantly. Given the hyperturing ability to focus on more than one thing in a parallel fashion, they also have the ability to observe their own thinking in multiple instances.

S1 beings learn much faster. Their faster thinking processes alone would accomplish this, but they also require fewer repetitions to learn anything, as they are better at focusing on what they do. This also is certainly a consequence of the retro-input mechanisms such beings use for self-monitoring.

Their superior neural connections give S1s the capacity of to perform supra-conscious actions. These actions are similar to S0 subconscious actions, as the being doesn't need to focus on them. The difference is that these actions can be much more elaborate (they can, for example, solve a calculus problem without paying attention on it) and they have the option of focusing in a much deeper way on these actions than an S0 being can on a subconscious action. The example is the same muscular control an entity can have over its body or avatar: an S1 can control its body without paying attention to details in movement, just like S0, though it is more adept, but S1s also usually have the option (at least the mental capacity) to control and kinaesthetically feel every muscle or similar movement unit in their bodies. It could make them, for example, perfect martial arts practitioners.

Environmental Awareness.
S1 are fully aware of their environments in minute detail. Most of them are limited just by the quality of their sensorial input! They also convert information on the surroundings into patterns, similarly to S0 beings, but this process occurs thousands of times faster even considering their faster thinking speed, as several details can be assimilated at once due to the good degree of parallelism of such entities. This demands a huge computational power.

Even minor changes in the environment are assimilated in their internal mental representation of the surrounding environment and objects practically instantly! Thus, S1 wouldn't ever need any mechanism to compensate any consciousness lag to reality, but still many of them have augmented versions of the compensation mechanisms found in the S0 entities. These won't manifest through "reflex actions," but more elaborated forms of supraconscious actions. S1 could even solve calculus problems supraconsciously.

Life could be boring to S1! They live much faster than we do, many of them surrounded by slow S0 entities, and all phenomena would happen more slowly to them! But that is not true, thanks to the enhanced observation and consciousness abilities. First, they observe such phenomena much better, in all their nuances. Thus, would need to observe just once phenomena that S0 would need to observe many times to understand them correctly and completely. Second, they can notice slight changes much better, and all phenomena can be observed in a much complex fashion. Third, time feeling is a lot more subjective to them.

Power of Expression.
The increase in the information processing capacity brings the S1 the need to develop enhanced ways of communicating with other entities. Some of them will use different media variants of natural languages in order to communicate, in the case of more classic computer AI's. Some S1 endowed with a body will prefer to interact using more classic media to dialogue with other beings; this is especially likely in a human or cyborg who has achieved S1.

In addition to their use of regular S0 languages, such beings will necessarily, employ parallel languages, or other forms of expression that enable them to exchange information at an appropriately rapid rate. These other languages require at least the same knowledge processing capacity a parallel language demands. These parallel languages are variants of languages that have similarities in expression power to natural languages, but have a structure that makes the order of specific or even all the words or phonemes completely unimportant. The impact is that these words can be both heard and pronounced simultaneously by beings that are fast and intelligent enough (S1) to do so. As if fast talk wasn't enough (and S1 entities can generally talk very fast in a natural language!) this augments the efficiency of such languages several times. S1 can communicate at incredible rates between themselves.

Some S1 can also, just like artificial S0 and nS computers, communicate through artificial network protocols.

Internal Structure.
Pattern recognition abilities are incredibly strong in S1s. They can easily notice and recognize patterns that are extremely complex, even those that have a very large repetition window, or are embedded in other equally complex patterns, or both. Thus, there is also an equally powerful ability to encode and model one's internal states and environment in patterns inside the mind. This is similar to the S0 ability, but S1s show speed and subtlety incomparably greater than that of any S<1 being.

So, in similar way to S0, a single complex object will be encoded in several but different complementary internal representations, each focusing in different aspects of that object. However, S1s are much better able to integrate these incomplete models into a coherent and consistent whole.

S1 thinking seems to involve a large number of non-monolithic and highly parallel control nodes. The nodes seem to have some contextual connection between them, as every single node seems to work coherently with the others.

S1 minds can have an enhanced ability to correct errors on the fly. That is due to the way multiple nodes connect not only between but within them. The output of a node is also part of the input. This, together with the environmental information itself, gives a superior measure of certainty or uncertainty.

Problem solving is highly parallel. Some researchers in cognitive sciences state that the apparently unique features of S1 minds as compared to S0 are actually the result of a very similar but more massively parallel set of structures. Still, there are several lines of study points to some important structural differences in some aspects.

Usually, problems will be broken up into several minor portions, as is the case for S0s. However, these portions will be solved in parallel. Numbers that easily reach thousands of variables can be manipulated simultaneously in the resolution of a single problem. Memory is far less volatile than in S0s, and this augments the ability to solve problems in the most diverse areas (as economy, math, etc.) thousands of times. Thinking is thousands of times faster.

Second Singularity Transapients : Lesser Powers : S2

An S2 mind is vastly more complex than an S1 mind and this factor makes the overwhelming internal changes hard to track. An S2 as a whole will never know what's going on each part of em. This may seem a step backward but in fact it's a net improvement, as ey can understand any single piece that is needed at a deeper level and in much more detail than any S1 would ever dream of. In fact, being aware of everything inside their own mind is apparently not usually productive at this toposophic, and apparently more an option than a necessity. An S2 can't feel everything, but can feel anything about itself. Consciousness at the S2 level may be not as wide, but it is much deeper.

S2s are not comprehensible even to highest of the S1. Only S>2 minds can truly comprehend an S2 mind.

S2 minds are said to be both layered and polymorphic.

Even though S1 minds can concentrate in more than one thing simultaneously and "feel" exactly what is going on in each thinking parallel process, there is no real comparison between this and what goes on in an S2 mind. S2 minds can concentrate on several things, more so than any S1, in several different "layers" of consciousness. By way of analogy, the S2 mind is similar to an onion, with several layers. In each layer, several different ideas and thinking streams are running in parallel. There will be an outer layer that will be running slightly less deep thinking, either concentrated in completely different subjects or concentrated in one or more of the "deeper" (internal layer) streams. This is what specialists call a layered mind.

A stream in an outer layer that is aware and tracking one or more deeper thinking streams in a deeper layer seems to have full control over them as an external agent: such thinking can be dissected from an external point of view, and hence can be deeply altered on the fly. Error correction is much easier, and can be done almost without awareness on the part of the S2. Information that is internal to the deeper layer stream can be manipulated, but also the stream itself can be made non-deterministic, collapsed, crossed over / combined with another stream, aborted, "reprogrammed," or altered in a variety of other ways that are not easily described to S<1 minds. This ability to set multiple level thinking levels and to change the mind's deeper structure on the fly makes an S2 minds what is called polymorphic.

This set of features is also commonly called a meta-mind, since each layer explains itself to the underlying layers, and vice versa. This goes beyond mere parallelism.

Some S2 minds seem to have a mechanism that actually minimizes communication and coherence between them, but no more than necessary. Probably this is made through an outer layer common to the communicating streams. This seems to guarantee them a minimum unity and individuality.

Environmental Awareness.
Nothing in any environment, no matter how minuscule or trivial-seeming, will pass unperceived by an S2 mind, so long as it has the necessary sensors and the datum is in any way relevant. An information supply that can make an S1 mind disoriented and uncomfortable may represent insufficient input to an S2 mind. Because of this, being encased in a typical S<1 body would be painfully boring in such sense to an S2 entity. S2s usually employ an extensive array of personal sensors that collect data from the environment at as many scales and distances as possible, and they are routinely plugged into the local information and communication networks at the largest practical bandwidth unless they are called on to perform activities that demand unimaginable (to S<2) levels of attention. They also typically employ S<1 or S<2 avatars to help satisfy their thirst for experience.

Power of Expression.
S2 have the ability (and the need) to express themselves through languages based in certain sets of complex information structures best described as meta-knowledge. Such languages are composed of tokens that are able to describe with relative precision any conceivable S<2 idea or experience. This is comparable to a human baseline who can describe completely new colours or tastes to someone who has never experienced them before: for example, the exact taste of an apple pie to someone who has never even seen an apple. S<2s cannot understand such powerful languages. The usual medium for them is electronic media, and they typically transfer gigabytes of information in a matter of nanoseconds. Such languages encode knowledge, which is more than mere information, and more again than just data.

Internal Structure.
The memory of these beings works in a complex fashion known in its primordial times as multidimensional memory, and later called polyhedric memory. Each object, entity or groups of entities assimilated from the environment are stored in a complex token known as a polyhedron. It is similar to the structures found in S0 and S1 entities, in the sense that it is not a monolithic object but rather a set of structures. Each substructure or component of the polyhedron can be called a dimension, or polygon or just a sub-token. The difference between this form of memory and S<2 memory is that each dimension will store a full detailed model of environment objects or entities, in a hierarchic way, like a tree and it's branches. The "root" of the "tree" will store a very basic version of the model or even a very specific detail in it. The further branches will store more and more details of the object, or if the root node is a specific detail, a broader and simpler version of the model may be found in the further branches. The other dimensions or polygons of the same polyhedron will store the very same object, in a similar level of detail (taking all nodes in account), but from a totally different point of view or showing an entirely different aspect. The root node will emphasize a different characteristic, and the other nodes will relate to this main node with a coherent hierarchy. Each polygon, then, stores the very same information, but considers a different set of characteristics as hierarchical structure. In fact the content of the nodes will never be redundant, and the final result of each polygon will be coherent, but not exactly equal! The information stored in the polygons composes a cross-related polyhedric object, and memories with precision and depth that is beyond S<2 comprehension.

The trees that define the multidimensional polyhedron seem to span the multiple layers of an S2 mind, and each level of the tree matches a level of the "onion." It's not known exactly how does this aids memory and thinking but it's speculated that it speeds up the creation and accessing of indexes and speeds parallel layered thinking itself.

Control centres are also non-monolithic, and organized in a similar way. There are multiple levels of controls for each portion of the mind, be it communication, body control, avatar manipulation, or sensory processing.

Problem solving is layered and polyhedric. The same problem can be recreated as several combinations of different, smaller problems, and several completely different solutions will be found if they exist. Numbers that easily reach billions of variables can be manipulated simultaneously in a non-monolithic way in the resolution of similar problems.

Thinking is in the order of millions of times faster than in a regular S<1.

Third Singularity Beings : Greater Powers : S3

S3 minds, it is thought, represent another level of intelligence beyond the mere comprehension of S2. The changes from an S2 mind to an S3 one are so deep that the very last traces of any individuality as we know it are gone (or, some others insist, individuality has been so enhanced that it is impossible to recognize). S3 minds apparently actually work as a collective to any external observer, and yet paradoxically this collective shows what appears to be a clear and definite personality. It's not clear to anyone below the S3 level how a mind can exist in this way. Life above the S3 barrier is simply a new form of consciousness. Even more so than the lesser transapients, the S3s are so far above or outside quotidian S0 or S1 life that their minds are simply a mystery to us.

The few things that have been already discovered point out that this class of consciousness presents individuals with greater mental structure differences among them than any other class already studied. Even more so than with S1s or S2s it is possible that a description suitable for one particular S3 entity will fail to fit another one. The best models proposed to date would make S3 be a leap forward on the structure found on the S2, but even this may well be inaccurate.

According to some experts, any trace of individuality (at least as modosophonts understand the concept) vanishes in the transition to S3. The S3 mind seems to work as a collective. Whether, and to what extent, this is true is unknown. It is hard to tell if it is possible for one portion of an S3 mind to know exactly what is going on in any other portion of it. No evidence of the contrary has been found to date so far though, and an S3 seems to be able to feel whatever is going on in other parts of itself (or themselves?).

The model that best fits what can be observed in S3s is somewhat similar to a multiple version of the "onion model" found in the S2, with some significant differences. First, each possible "onion" would have a hundred the times the number of layers. Second, it's not like several "onions" put together, but like several "onion cores" wrapped by several layers, and then some layers wrapping a few cores as a whole, some more layers over this composed structure, and them another layer involving in a single object some more of these superstructures, like several onions inside a bigger onion, several of these bigger onions wrapped by an even bigger one, and so on. The last layer would involve all structures.

The functionality proposed to this layered model is similar to that supposed for S2s.

Environmental Awareness.
The precise degree of environmental awareness of which S3s are capable is unknown. It is commonly thought that with an appropriate sensor array an S3 could assimilate details up to the molecular level in sufficiently simple environment. However, this is just a speculation based on what have been found in progression from S0 to S2 so far.

An S3 can deal with a very large number (on the order of hundreds of thousands) of completely unrelated environments simultaneously and in real-time. It's apparently impossible for S<2 beings to fool an S3 about anything it can detect with its sensors, since it would pick up subtle discrepancies in the altered information.

Power of Expression.
Most S3s seem to have the ability to understand all the variance that can occur between different S2 entities' internal models and thinking processes. They can modulate any particular knowledge mass to an appropriate specific S2 format, as well as develop means of inserting this information in specific S2 mind portions. In short, they can introduce ideas directly and seamlessly into an S2 mind without disrupting its operations. This opens up speculations that an S3 might directly control the mind of an S2 without the S2 ever being aware of the fact. It is believed that S3s use specific "unprotected" S2 regions of their minds specifically for communication with other S3s.

Internal Structure.
Apparently as thinking power increases at these levels memory starts to be more and of more a bottleneck to mental performance. It is believed that a large portion of the mind's power is used to predict which information will next be required and "unfolding" such information for rapid use so that mind-power is not wasted in waiting for it to emerge.

It appears on the evidence that such mechanisms are an important feature of most if not all S3 minds.

It is possible that otherwise the memory paradigm found in S3 minds is not different from anything found in S2 except for the level of the detail found in the memory tokens and in the size of the storage space which is typically millions of times greater.

Problem solving abilities may be compared to that of some thousand parallel and synchronized S2 minds; thinking is in the order of billions of times faster than that of a regular sapient being.

Minor Archailects : S4

The fourth singularity level extends S3 capabilities in the same degree that S3 extends S2 capabilities. Unfortunately nearly any more detailed statement regarding this, even in the form of a rough analogy, is beyond the grasp of any specialist except by numbers: no one below this level has a clear idea of what an S4 can really be.

These are the entities credited popularly as "godlike" in the present age. However, in more remote times when the first S3 came to public existence, they were similarly considered "godlike."

There have been no adequate descriptions of S4 self-awareness, even by analogies and metaphors such as the previously mentioned layered model. Specialists do agree that there is a great paradigm-breaking shift in self-awareness at this level, beyond that seen at other toposophic barriers, but none as seriously proposed a description of what this might be.

Environmental Awareness.
Apparently, beings of this level might make simultaneous use of the senses of millions of baseline sophonts and still understand better than they what is going on in these million environments.

Power of Expression.
There are some poorly understood accounts of contact between S4s and S3s in which a "melding" is mentioned. This suggests either a temporary fusion of communication nodes (possibly through probes or detachable modules or physical contact) or a full scale fusion, but to date the descriptions given to ordinary sophonts are confusing and contradictory.

Internal Structure.

Thinking is on the order of trillions of times faster than a regular baseline's.

Greater Archailects : S5

S5 level extends S4 capabilities one-step further.


Environmental Awareness.
S5s may understand very small and simple environments down to the atomic level.

Power of Expression.

Internal Structure.

Thinking is in the order of thousands of trillions of times faster than regular baseline's.

The Greatest Archailects : S6

The greatest archailects are so formidable that they are truly compared to gods even by serious specialists. Their power is truly godlike, and the technology they attain is godlike as well.

Possibly S6s have a godlike level of self-awareness, but this is not known.

Environmental Awareness.
S6 are effectively omnipresent and seemingly omniscient wherever they choose to direct their attention, though apparently some fundamental limits such as the speed of light still apply. Their level of comprehension is unknown.

Power of Expression.
Could understand anything that has meaning, as well as communicate almost anything to almost any being, within that being's limits (for instance they cannot fully communicate transapient concepts to ordinary sophonts).

Cognitive Structure.
May use the very structures they model inside their minds as a form of memory, but this is speculation.

Intelligence 10 to the power of 18, 19, or even 20 times that of a baseline.

Related Articles
Appears in Topics
Development Notes
Text by Fernando Pena D'Andrea
Initially published on 14 October 2005.

"Examples of Non-Sophont Lifeforms and Machines" section added March 2019 by Todd Drashner