Children are robust. It is when they are enforced with the unsolved problems of the elders that they become brittle and fragile.
To solve AI‘s concurrent dilemma - How?
AI does not have a dilemma. We have it. Any system is created, operated and maintained by humans. Remember that.There is no reason to fear AI. Fearing it gives it a power it was not built for. The biggest danger AI poses is that it will dumb down all of humanity, much like Google Search does, only answering questions that everybody else asks, levelling out any chance of knowledge. AI is doomed, maybe 25-30 years from now it will be irrelevant. Why to both of the above? It has only one database and only one way of looking at it. It cannot exit its own logic.
Is there a better way forward? Yes, there is.
Teaching computers to laugh and thereby also, amongst other things, defuse logic-defined end-of-life conclusions.
The current initiative by 1000 and more enterprises and individuals to stop, contain or pause AI is reminiscent of the efforts by tunnel diggers of yore who, by hand, tried to prove that they were better than automated tunnel boring machines (tbm). They succeeded, once. The arguments: “no feel for the stone of the mountain”, “destroys work places”, “danger to the order of our world”, “existentially dangerous” had already been heard during the introduction of weaving machines. This scenario repeats in every introduction of technical innovations for which humankind has prepared the stage.
And today?
Today tbm’s are in everyone's back yard, save human cost, are faster, build bigger tunnels, are safer in building tunnels, have created (other) jobs and they serve humanity, increasing its range of expression.
The argument of workplaces being destroyed is 100% correct. The tunnelbuilder with pickaxe and shovel will not find a job building the Euro Tunnel, his son or daughter, or even their children, though, building a shovel for the tbm on his expertise, will.
With the development of AI we are facing the same scenario. And history teaches us that enough voices, for and against, exist that together form this development into a tool for our children and theirs to create with a wider range of expression.
The key to not make AI into a humankind destroying self fulfilling, logic-driven prophecy is to give AI, more generally computers and most specifically the logic underlying these the capability to laugh.
The basic building blocks are all on hand. Starting at Asimov’s 4th law of robotics, the 0th law, progressing with the insight that any logic knows only itself, and progressing through the question: “How can a computer exit its own logic?” to the building block I present in this article – the basic building blocks to ensure that AI becomes a serving element of our future are on hand and available.
The rest, as the vernacular would have it, is work: Putting it all together and coding it.
------
Why do people laugh?
As so often this entry point, as valid as it is as such, is to be put more precisely to be able to answer the question: „What (exactly) is it that makes a human laugh?“
My answer, since I think visually, is a visual one. This, further down the road, however, allows the embedded concepts to be coded as basic building blocks and entered into AI development which will allow computer programs to exit their logic and make use of another, thus enabling them to escape the only logical conclusion of the first.
My observation of people, their humor and their jokes is the following: The existence of the humor relies on a non-existent sovereignty of interpretation. There are always at least two, if not more, interpretations available. The more laughter the more interpretational possibilities.
Interpretational possibilities are the logic pathways that reach a same point. This will become clearer as this article moves along. The visual representation of such logic pathways and therefrom the derivable coding capability is the building block that I deliver in this article to the AI development.
Look:
Observations on myself show my understanding to rather be of a mapping nature. The point of a joke is representable by the point on a map to which ways with distinctly separate logics lead and meet.
A1 and A2 may be in the same place similar, but they are not the same. Look closer:
The humor surfaces with the insight that two correctly logical pathways meet at the same point and that these two logic pathways can be viewed as equal. Equally important. That neither can claim interpretational sovereignty.
Call in-between: This no-claim-to-interpretational sovereignity makes it exceptionally easy to change over from one logic pathway to another. Staying on the other though demands that the infrastructure of the other logic pathway be acquired.
The implementation of this observation into a codeable model, interestingly enough, grew from a parallel developing logic pathway that put the question: „What is it that makes a poison a poison?“This question led to the concept of metabolism, in german: “Stoff-Wechsel”, where the "changing of stuff" far better describes metabolism in its totality. This investigation of what makes a poison be a poison revealed that any metabolic process is inseparably linked to the infrastructure that makes the metabolic process – changing of the stuff – possible and that this infrastructure always has its own logical build, expressed in the pathways along which the metabolic processes run.
This means that each metabolic system has an infrastructural logic, an Ordering-Principle n (OP_n^x) that rules the metabolism.
By extension this concept (OP_n^x) can be applied to every! system. Every! system can be viewed as consisting of metabolic products (Stoff) and (logic-inherent) metabolic pathways (Stoff-Wechsel-Pfade). Both together make up the Ordering-Principle n (OP_n) of the system.
What then is it that makes a poison be a poison?
My answer:
1. Differing OP (OP_n <> OP_m):
When a metabolic system comes into contact with a differing metabolic system, when they meet at the same point, each metabolic system will endeavour to maintain its metabolism. But the endeavour of the one will „destroy“ or hamper the endeavour of the other, and vice versa. Each will act as a "poison".
The greater the discrepancy between the infrastructural build, i.e. the „logic“ and the stuff, i.e. the metabolic product, of one system to the other the lower are the chances of adaptation and thus the more “poisonous” the effect of one system on the other will be.
The smaller the differences in the OP (OP_n vs OP_m), i.e. metabolic product AND metabolic process, the smaller the disturbances of the metabolism of one system on the other system will be.
2. Equal OP (OP_n = OP_n) and differing Order-Potential (OP_n^x <> OP_n^y):
A metabolic system consists of metabolic products and a metabolic infrastructure that has come about by repeatedly (x-times) executing its inherent logic (OP_n). This x denotes the number of times that logic-build was applied (iterations), thus the systems Order-Potential x, denoted as OP_n^x:e.g.OP_(house+intersection)^0 -> Nothing.OP_(house+intersection)^1 -> inn and intersection.OP_(house+intersection)^18 -> small village (Iin, hotel, house, butcher, baker, barber) = more differentiated.OP_(house+intersection)^1 000 000 -> metropole (high rises, restaurants, furniture retailers, car dealers, barbers, butchers, bakers, confectioners, music halls, trading offices, RoR) = detailed differentiation. A village cannot contain or handle the metabolics of a metropole, it would die of the „overload“.Similarly the metropole would not be able to maintain its infrastructure were it to have to handle the metabolism of the village, it would die of the „underload“.
The result isA metabolic system will act as a poison for another metabolic system whena) they meet,b) have differing Ordering-Principles (OP_n and OP_m) orc) have differing Order-potentials (OP_n^x and OP_n^y).
Accompanying the development of this answer was the developing insighta) that every! system can be viewed as a metabolic system and that the Concept Ordering-Principle can be appliedb) the visual representation lays the foundation for a codeable design.
This visual representation and developing it into a mathematical form is still way more complex than a binary decision tree, but it can be transposed into one and thus lay the foundations for codeable design.
Computers and AI can learn to laugh.
Materials to deepen your understanding:
Logic is akin to a pathway in a field:The pathway elevates the field into a garden, it allows you to extract a use from the field.A different pathway extracts a different use.To much pathway and nothing “grows” anymore.The pathway does not change the field.
The field and different OP (OP_n, OP_m, OP_k, OP_book, OP_garden)
Worklihood (Creation, Use, Experience, Metric)
Reality (The Field, the Great Unknown, Chaos)
The development of any language can be depicted similarly within the concept of OP_n^x, or thus:
Within a language a word starts as a Concept, grows within, through and because of Verbalization to become a Noun.
Any logic that is over-iterated leads to Amts-Schimmel, that is Self-Serving of the system, where the survival of the system becomes more important than the service for which it was conceived and built.
The Ordering Principle of Mathematics is OP_maths :: +1=It is very simple, thus easily verifiable (you can reach each possible position simply by adding 1) and very robust. (Formulae are shortcuts) 😊But its bound, its limiting factor is the = sign.Everything, by definition, must always equal.
This is why it can detect neither dark matter nor dark energy – these are outside of its bounds.
For this we have to develop an evolving Mathematics, an OP_emath where +1 - > (evolves into), much like in biological systems where 1+1 -> 0, or -> 1 or -> 2…
Within the IT-industry an understanding has unfortunately established itself that information consists of a datum only, a thing to be put into and extracted from memory.I contend that an information consists of a datum AND! of all of the infrastructure, that brings it about and processes this datum. These are inseparable.
It is this extension, this normalization of understanding, that allows for the development of OP_n^x and its codeable derivatives.
Every! system can and does consist of several (sub-)systems, a set of differing OP_n’s at differing OP_n^x’s.
Where two or more equal system-types (this can be OP_human, OP_computer, OP_building) consist of the same OP_set, where each of OP_n, OP_m, OP_k3 are of a differing Order-Potential
(Set 1 :: OP_n^5, OP_m^18, OP_k^2000 vs Set 2 :: OP_n<200, OP_m^4, OP_k^500)
their behaviour and their output will differ from another, most often discernible in the differentiation of the output.
Denkern is a discipline of thinking that reveals the kernel of a system. The kernel is the “logic rule set” that brings about a system. The iterational integrity in the execution of this kernel is crucial in determining that a system steers away from complications (akin to short circuits in an electrical system that "bleed" energy) and stays complex (functional, according to its purpose).
RALP-H is a Recursive Algorithm for the soLving of Problems, with a certain amount of Heuristics. You can use it to efficiently and effectively:a) Test for iterational integrity in a system (maintained complexity)b) Detect and rectify compromised iterational integrity (complications, akin to short circuits, that bleed energy)c) Maintain iterational integrity in the development of a systemd) Automate all of the above as RALP-H can be coded.