The belief of artificial intelligence designs
For much of the history, as well as the current conception of artificial intelligence designing, the design and implementation of AI has been influenced into the four main paradigm, as mentioned in Norvig, Artificial Intelligence. A modern approach (2011). Alternatively, it also follows the more general view toward either looking specifically at the structural components of any specimen of intelligence (human neurons and animals’), or neurological approach, also called connectionism; and those that follow the view of universal rationalism, capturing the essence behavioural constructs, and basing AI constructions on such universal justification, the symbolic AI approach. While those theories can be said to eventually be useful and perhaps, more groundbreaking than not, they are not so much desirable, as they are also very well-being specialized, unscalable, and was stuck in the paradigm of the more simpler tasks. In fact, whilst NLP, the chatbot that is prominent in the later date of the 2010s was hailed to be the epitome of development in AI (which indeed, perhaps it is), the issues with, for example, common sense knowledge is still present, and any attempt in fixing it often resulted in either the rough, uncertain and naive mimic of the ‘real thing’, or fails miserably. In one way or another of such, we can say that the current AI theory has no framework. Or at least a manageable one.
It is then of our interest to put up at least a view of what should be, and what can be said of artificial intelligence as a framework, a system for theories, and the principles in which one can say about the core essence of artificial intelligence, given an interpretation. In said following sections, we would allow ourselves to tread the thin line, and lay out the foundation of what to develop to a full-fledged artificial intelligence theory.
What should be expected of this outline? Not so much. The ultimate goal here is to formalize the principles, the working mechanism, the designing compass in which developments might ensue of a later date. Settle on matters that of said digression, will influence how the entire framework is formed, from the macroscopic view to the microscopic design elements, the point of view of the system, and more on the treatment of such system in certainty. Or probabilistically, which you can choose. In fact, the issues between determinism and uncertainty (probability interpretation) will also be touched upon, forgoes the huge amounts of works on two said interpretations ambiguous as it is possible to be.
I. The definition for artificial intelligence
There are many ways to define artificial intelligence, either by the phenomenological one, or by the presumptuous, universal-assumed way of logicians in the old way. However, our definition, or at least mechanism of working on artificial intelligence, should be by then very different.
What is then called AI, in our view? We treat AI as rather a more general system. First, we digress on the term artificial. We have talked about this, but, for the moment, it is perhaps better to reformulate it. A construct, object, subject \(A\) is called artificial if:
Definition 1.1. (Artificial)
We say an object \(A\) is artificial only if it is not natural, or rather, it is intentionally created by meaning intentions, and not the general evolution of states, the natural interaction of the law of nature, or the natural transformation of biology.
By such definition, artificial is then a quality to be separated from natural intelligence - of naturally made intelligent vessels or construct, existed because of natural, biological evolution and advancements by itself. A construct, then, is the main interest of our study. What about intelligence then, one might ask? The truth is, we don’t know. We know roughly that intelligence is the exhibition of rationalism, the actions and demonstration of perhaps consciousness, of thinking. However, we have no way to define operationally, formally, and if not, conceptually sound of such aspect of intelligence. Let’s assume we don’t know. Then artificial intelligence is especially the realization and the discovery of intelligence itself. We have the following definition to perhaps partially reflects that, or rather aligns well with the general perception of what constitute the choice of the words artificial intelligence.
Definition 1.2. (Artificial intelligence)
A construct \(U\), subjected to a system \(S\) is called artificial intelligence if it satisfies the condition of being artificial, whilst also satisfies a given criterion set of being autonomous, dynamic, and overall general. It is such that will give rise to intelligence construct.
This is a definition that encapsulate quite a lot of way that we usually conduct on ourselves, when talking about artificial intelligence. For once, we recreate it using some basis, or construct. Then, we make its functionality to work according to certain criteria that we set ourselves, and if it satisfies such, then it is called intelligent. Specifically, one of such tests are the test of rational thoughts and logics.
We will further develop those points made above. However, we might as well want to justify the first point in all - the point of the intelligence criterion, and why we might not need it.
I.1. Criterion of intelligence
We mentioned the notion of the criterion of intelligence. However, what should we define it? How should we know to even evaluate it, is a very hard question even that we did not (or unable to) fully realize yet, then what we want to do with it? This question is where a lot of things in the artificial intelligence research was based upon. For example, the (Total) Turing Test in which outlines possible outlook for intelligence, for capabilities that then defines the fields in which we are having nowadays, for example, computer vision for the capability of visual perception, natural language processing (NLP) for the capacity of language, and more. We also have various conceptual criterions in which people have been suggesting about the model of the intelligent being, for example, various set of criterions that outlines and includes even consciousness, some suggest behavioural conditions, some goes for the exhibition of chain of thoughts, and some even goes further than that, which is perhaps irrelevant aside from mentioned for example. Overall, it is perhaps a mess.
We still do not know what to come of criteria, or rather, in the quest of producing intelligence, we base ourselves onto it too much. As a species capable of intelligence and more sophisticated notion, we have the basis, and the advantage of being able to examine ourselves. By that, eventually, as the highest example of intelligent being, we use ourselves as standard, for examine, psychology, neurological behaviourism, neuroscience, applied onto the quest of going for artificial intelligence. Hence, there exists the total Turing test, and there exists the conflicts between various definitions and criterion of artificial intelligence. A mistake perhaps has been made, doubtfully so that one did not realize of such. While it is said that AI researcher has been working on, or at least researching on the general notion of artificial intelligence principles, it is, in fact, not so much of a principle, as we did not realize yet that what we are doing is still the act of mimicking ourselves - creating a plane by replicating a bird. By phenomenologically absorb and construct architectures, models on the higher-level surface of what artificial intelligence constitute, the deeper construct is still non-existent. By copying the apparent capabilities of human and related intelligence being, biological rather than not, the core of which those behaviours occur, and facilitate the organs and observations made is perhaps, manifested. Ironically, while being too strict, wrongfully abhorrent to the fallacy of themselves, and too resistant to changes, symbolic approach got one of the right thing. If there exists intelligence, then it must be universal by virtue. That is, you cannot argue that alien from another universe is not intelligent, because they do not satisfy one of the criteria of the Turing test, just because such notion does not exist in such universe.
I.2. The should not of defining AI
Personally, I don’t think we should, or we could define artificial intelligence, at least of this particular stage that we are in. Philosophically, being an armchair philosopher would not help in pursuing such notion, yet again because we are arguing on the basis of our own existence, and not the subject’s matter viewpoint. There are problems related to it, also, of such that the mind and consciousness is arguably debatable in every given sense, of which no one seems to agree on the mundane notion that intelligence and consciousness come from chemical and the weird ‘quantum effect’ that would be then believed to be. And, truth to be told, we are not even endorsing such direction. In actuality, we don’t even know what is intelligent, and also don’t even know what can be of artificially made rather than matching mathematics.
On the flip side, computationally and neuroscientifically, the lack of formal treatment and overall encompassing knowledge conjunctions plague the construction and foremost attempt to do anything, simply because too many things have been said yet none can unify them together. Such is also to say different directions and different methodologies being conducted, yet they are so distinctively separated to be unable to conform one to another, despite them taking on the same object. Furthermore, there are a lot of assumptions given in computational theory, and the overall application thereof. As for anything, assumptions can be broken, and reinforced, for whatever it is being inconsistent as a virtue.
It is wise to remember that, for now with neuroscience being not advanced enough and in a perhaps different direction from what can be seen, while certainly for empirical science we can utilize neuroscience’s knowledge, we should not take in the philosophical arguments and ‘idea’, including computational theory of mind. For empirical neuroscience, it is also not the fully-encompassing field that observe the brain from every angle, and observe consciousness of everything if ever, at least of the present. And, for the philosophical and idealistic view, only one thing can be said about such being “the lines on the map is made up”.
Every ‘theory’ makes up the line on the map to draw out the world; yet the world might not work that way or might work the other way or whatever. Such is to say that to define artificial intelligence is then perhaps, not so much mingle of a fruitless endeavour for anyone to posit.
II. Percolation, Emergence, and Determinism
As we have been saying all around, one particular aspect remains. It is the discussion of the old time, of whether it is good for us to “construct the intelligent by virtue of its appearance”, or by “meticulously engineer it”. Such is to say, are we willing to give up understanding, in a given sense, toward the potentially feasible appearance of such, or handling the devilish hardness in determining everything of will, engineer it from scratch, of specific details? This is the context of emergence and determinism, in the sense of artificial intelligence ‘wishful thinking’. They have the merits, somewhat similar to the analogue to physics where statistical physics was developed in contest of the overwhelmingly complex nature of what constitute the overbearing system, for example, of ten thousands particles. No one would be willing to put up such analysis that would worth of a few million pages. Determinism offer understandability, which is of dire needs in many places, many applications, many contexts, yet seemingly ineffective of larger system, or because we do not know how to do it.
Of the two (or three, depending on the choice of words), emergence has become a buzzword for people out wide, of the feverish believers in AGI (Artificial General Intelligence) - certain thing that keeps changing its definition every decade or so - or of the optimistic camp of which believes in outright advance without looking back - giving up understandability, theory, analysis for the progress toward - whether if that is inching from 95.1% to 95.9% every conference. It is also of the attention of people who are of immense confidence, and of somehow ‘resolute knowledge privilege’ to other about life, meaning of life, the source of intelligence, god, the world, and whatever there are to discuss that either outright cultist stuff, maybe borderline pseudoscience, or simply just pseudoscience. The fact and the potential of ‘giving up to faith’ and randomness, to somehow gain AGI from nowhere, to somehow without that many efforts, simply get a hold of the concept that even by now no one understand, is perhaps one of the many wishes that people get, and thus attracting such attention (even in academic society). That is why we also often see people claiming we are 2 months away from having ASI (Artificial Super Intelligence), which is a, well, crazier version of AGI in the first place.
Personally, I agree with the philosophy of emergence, somewhat, just as I have discussed above. However, for every knife being double-edged (of course this is metaphor, don’t bring up one-edge knife here), the concept is itself having the downside that perhaps people neglected. I will leave it to you though. Speaking them out is troublesome, of the very least.
Nevertheless, it is of foremost interest to also realize that those that we criticized, did bring up advancements and expansion of capabilities and theories that guarantee ourselves of being able to criticize it. We have come this far ahead because of it. But no one said that we have to stick with it till the bitter end.
III. A more principled approach
So, how should we approach this particular problem where one wants to create more than just logic in disguise, but also computational in nature, or else that no one can predict? Well, it is to generalize them . Simply speak, we do not go for the AI itself, but what can then constitute it. Granted, it is not similar to going blindfolded, or any kind of predisposition that protest the usage of the term and outlook on AI simply because perhaps of the impression above that we do not know what it is, hence no need for pursuing on such narrow road. But rather, to extract the fundamental facilities, concept, objects that in conjunction of knowledge that can be brought up or newly constructed, that is relevant of interest. And starting from ground zero with the modesty of assuming none and bias to minimum.
So, what can be said of such so that we can continue working on it? Nominally:
- We do not create artificial intelligence. We create and investigate the facilities, the theory of which structures and objects can be utilized to create the framework that is not intelligent, yet encompass more and might be able to give rise to intelligence. That is, for example, if we are to say that artificial intelligence can be constructed on a machine, then what can be said of the theory of machine? What is machine? What is the model for machine, and our understanding of such? What are machines that do not bear any similarity to the skewed vision of what is intelligent? And so is computer and its principle, and how we actually, formally, operate it? More so, we construct the subjects. The intelligent comes afterward. 1
- We do not, consequently, gauge intelligence as per metric or in terms of the skewed test that one can take, or metric that one get the \(R\)-value and so forth. It is rather somewhat baseless in such regard.
- We assume the assumption of constructiveness - there exists the absolute minimum of any given object that can be the basis for more advanced concept and construction. Not regarding such constructiveness often gives us headache in co-joining different constructs and ideas together, or simply interpreting machines and constructions of their characteristics and properties rather than just ‘somewhat weird trick’.
- No model is perfect, if we ever create one.
- We regard artificial intelligence to be consisted of two main things: the facilities that support its existence, and the process that support the subject matter that is examined. By this, then, instead of for example, thinking that intelligence only exhibits in subjects that have the learning property. Then we go on the reverse. What would happen to any given subject, that has something even remotely similar to learning, without considering the case of “if it can learn then it is intelligent” and of what capability? That is, to consider the action to be components itself, and not ‘requirement’? Overall, without stating more for out-of-scope reason (I do not intend for this note to go further into such discussion), we are trying to construct and investigate the formal foundational knowledge first, before even can utilize it to construct a generalized construct that encompass what is not intelligent, and what is then intelligent. Then, construct it with assumptions and constructiveness, plus the realization of both the building, and the lifetime of such building by itself.
1 It also can be considered loosely as to focus on the chaotic behaviours and the role of percolation plus emergence, rather than, well, specific construction. But with a blend of both that and descriptive criteria, not relying on anything, would be of the best. Of course, what is even emergence in the first place? Are there definitions for it that do not require pub fights?
With this, perhaps we can then continue what was left behind, of what we might want to do and upgrade. If we are willing to put it of a test.