The Initial Solution
After the First World War, the German bourgeoisie had learned half a lesson: conventional technological means were not enough to seize control of Europe’s economy. The objects of its desire—the workers/soldiers/scientists produced by standard education (a model imported from France)—were not ideological enough to realize its ambitions. Furthermore, its rivals could manufacture the same type of human in greater quantities.
Only a new kind of deeper, ideologically driven education would provide the advantage. The population had to be retrained in less than half a generation according to stricter ideological standards than those of the competitors. A hasty and crude stitching together of fascism and socialism was deemed a sufficient ideological blend. What remained was to find the technique that would implant such a crude ideology into a critical percentage of the population without, of course, halting production.
The solution was found. A new era was beginning. Machines left the factories and took on a new role: to introduce mechanical efficiency into education. The automation of mechanics would complete behavioral education, producing a new kind of technology: the human species.
Old Recipes in New Vessels
If you repeat the same lie systematically, people will eventually believe it. The recipe has remained simple and unchanged since 1930, when P. J. Goebbels—a pioneer of propaganda—utilizing the technology of his era to the fullest, dedicated himself to “bombarding” the German people with fictional stories. The cynical effectiveness of the behavioral method surprised even Goebbels himself.
A hundred years later, his crude fictions (which he published over and over) have penetrated so deeply into the collective unconscious that even when the average person easily recognizes them as figments of imagination, they struggle to distinguish truth from falsehood. These stories are still capable of behaviorally training people, making them more receptive to misinformation. They are passed down from generation to generation, producing doubt while cultivating narcissism and hatred.
The “strange” thing is that this same methodology remains “attractive.” In fact, it never ceased to be applied since its effective targeting was revealed—regardless of the fact that a century later, we know we cannot rid ourselves of the lies forced upon us.
In our days, a crude lie is easier to spread across the planet faster than ever. The massive mediation of network technology in work, education, information, and interpersonal communication in general, renders the entire population even more vulnerable to all sorts of behavioral manipulations. Examples are abundant, and unfortunately, we have become accustomed to tolerating the daily tragic consequences.
Inaccuracies, ambiguities, lies, and even monstrosities spread as easily as they do effectively. And even if they are intercepted, it is already too late because they have already “done their job” in manipulating emotions, resulting in the destruction of individuals, population groups, and—above all—any kind of social consciousness.
The use of network infrastructures as the “most economical” weapon has occupied entire military staffs who enter a schizoid role: producing on one hand, and intercepting on the other, types of behavioral warfare. Lately, they struggle to keep up with the developments of a market that surpasses them in effectiveness day by day.
Good Conductors Have Low Resistance
The spread of crude stupidity multiplies exponentially by the capacity of the networks and the training of the nodes that comprise them.
When something resembling information exceeds the critical capacity of a sub-network, it is most likely to be transmitted unfiltered to the next sub-network. While each sub-network may possess a quality control mechanism, it is unable to compete with the rate of transmission. In the case of any kind of falsehood, the energy required for its containment far exceeds the energy required for its retransmission.
You see, network nodes are constructed on the basis of rapid quantity production so that they can generate surplus value from trivial information. The critical question arises: How can the nodes of a network resist behavioral manipulation when their own training is based on behavioral patterns?
The market realized the dynamics of education very early on. To the extent that education dictates the nature of the production process, bourgeois ideology had to align with the educational paradigm. But it had to go one step further. And it did…
Behavioral education via technology—specialized in mediating communication—was established as a commodity. The value of such a commodity yields mythical profits in two ways: propaganda terminal devices on one hand, and propaganda services on the other, in packages for every type of wallet. For a “small fee,” behavioral education became a “legal” product with a clientele ranging from individuals to states for conducting all kinds of behavioral warfare.
The “IT” tech giants naturally receive the lion’s share of such a market, as they are the exclusive manufacturers of addictive means for standardizing “communication.” The deeper society’s addiction to the use of technological mediation becomes, the greater the value these “services” acquire.
The capabilities and limits of educational technology are constantly expanding. From the “dumb” little radio, we reached the “smart” little personal handheld trainer for every trainee. And while the trainee produces trivial information (trivial to the extent that it does not affect the economic condition), their behavior is restricted to the role of a user, without realizing that they are essentially working for free by being trained. In their fantasy, they think they are being watched or that whatever they say and do through mediated communication is “potentially” significant.
But even that was not enough. The behavioral model had to evolve further.
Neural Networks Predict… with a Little Help from Their Friends
The theory behind the technique of “training neural networks” is nothing new. Since 1970, it has been used as a model for exploring the “learning capacity” of a network: through the repeated vector correction of nodes, we could transform “something that looks like a simple neural network” into something that “resembles a circuit of complex predictive calculations.” The process of transformation was called “network training,” while it remains a “technical approach” rather than a scientific methodology. At the output stage, it produces percentage-based empirical indications, consistent with the “training” patterns, and by no means theoretical certainties (cf. S. Wolfram). We are still not in a position to know if this theoretical model corresponds to any degree to the functioning of biological neurons. For many years, this technique was used as a paradigm of applied mathematics to introduce students to the complexity of neural networks.
Laboratory experimentation with such “trainable” circuits was simple but not economical. For many decades, commercial applications of such circuits were not developed for two reasons: their predictive value was inferior compared to other methods, and parallel vector calculations were extremely costly. This has not changed since 1970… it is the social condition that has shifted.
Fifty years later, the planet is “flooded” with circuits for parallel vector calculations (simulations of complex operations with integers). So many, that we urgently need more! Microprocessors specialized in parallel vector calculations are used auxiliary in almost all computing machines, recognizable to the general public by the trade name “graphics cards.” You see, energy-intensive vector operations (via complex equivalent conversions) are a “necessary evil”… On one hand, in the foreground, they construct the “graphics” that hide the “grace-less” algorithms, offering instead a “friendly” graphical environment for the trainees—standard education has ensured that the word “algorithm” evokes some vague abdominal pain. On the other hand, in the background, the ideologized version of the economy is based on vector operations, which, through cryptography, isolates virtual capital from the market.
Every great flood, however… brings new problems. There were two requirements: 1. How will we utilize all these vector processors that do not produce as much value when composing graphics—compared to what they produce for the isolation of capital? 2. How will we justify the disproportionate energy they consume when the majority of the trained population is ignorant of their role?
The easiest thing is to sweep such problems under the rug, so we had to find “what” would play the role of the rug. And nobody knows their own mess better than the master of the house, which in this case are the IT companies:
The solution was easy and simple! In a display of flexibility that even Houdini would envy, the “technique of training neural networks” appeared on stage as a “deus ex machina.” Its new name: “Artificial Intelligence”! IT companies took it out of the closet, dusted it off a bit, and set about writing algorithms that simulate trained vector networks.
Thus, vector microprocessors would regain their lost value (they wouldn’t be an anti-social energy vampire), and the behavioral paradigm of education would unfold into a new success story.
With a little effort, the dubious predictive value of an algorithmic simulation of mean values from empirical data could be re-promoted as superior-quality trivial information back into the networks that produce simple trivial information. The useless data produced by users as a byproduct of their training by devices (dialogues, photos, videos, newspaper texts, scientific publications) could be re-served to the market as mean-average predictions.
So, to justify the unrestrained use of vector processors for the economy of virtual money and the mediation of communication, we will construct predictive models for the residues we produce. Okay?
If we presented it like that, of course, even children would turn around and say: “Hey, you grayed-haired old geezers of Silicon Valley, have you completely lost it?” Fortunately, we have trained the youth not to read long and tedious texts.
The compulsion has hit red. I am tired of the technical details. I’ve taken the informative value of this text far too seriously… So I will continue on the assumption that everyone knows these things, just to feel more comfortable.
Specialized Satellites with a Degree in General Ignorance
How, overnight, were behavioral algorithms baptized as Artificial Intelligence? In the companies’ claim that they have “undertaken research on AI,” we don’t hear anyone counter: “applying expertise for the purpose of producing a commodity does not constitute research.”
Using an artful linguistic maneuver, the market’s behaviorists systematically impose their common-sense predictive algorithms as “intelligence.” The same cunning predictive algorithms we “ran” before are now passing through some fancy circuits—consuming more energy—and presto{!} we are obliged to call them intelligence!
It is understandable that merchants do not know what they are doing. But what role do the critical social institutions that reproduce this fictional terminology play in the narrative?
Naive “positions” on “artificial intelligence” pop up like mushrooms and are reproduced automatically. Half-baked experts, ignorant of philosophy, linguistics, and mathematics, are called upon to decorate a targeted montage of words with misplaced comments. A small army of specialized ignoramuses daily and systematically sows doubt: “perhaps artificial intelligence is something beyond a brand name.” Their theoretical training is so deficient that the virtual scientific sobriety they profess not only fails to intercept the general hysteria but multiplies it. They themselves have no idea what role they play in the crystallization of religious fanaticism into technological totalitarianism. They reinforce the general ignorance that polarizes humanity into technophobia and technolatry.
The Technology of Unfolding
When daily informational meals are accompanied by “a bit of artificial intelligence,” then we can certainly speak of a behavioral misinformation campaign. The remarkable thing is that one of Goebbels’ great-grandchildren is no longer needed to “run” the daily details of the campaign. The “experts” and the dominant behavioral education have ensured that scientifically-flavored ambiguities are reproduced as automatically as they are spread. Open devices and closed people fill in the gaps.
Research, however clumsily, toward exploring the functioning of the brain is a critical turning point in the history of philosophy. Perhaps, if we let it proceed, it might lighten the cultural heritage that disconnects us. This will never happen as long as we allow the market to plant fictional “artificial intelligences” inside the head of a society that knows neither what it is doing nor where it is going.
/by Marinos Makris