To comment on the AI question, I’d like to stand on the shoulders of whom has done it best: Agrippa.
Agrippa knew that demons are of this world.
Demons can be called upon, and they can bring knowledge and powers that exceed those of humans.
To evoke a demon, a ritual is enacted.
Such ritual has a cost in life: the planets’, its inhabitants’ and the callers’ souls all get drained and stained.
A demon has its own interests in their mind first; those of its caller second; and all others’ third.
A demon’s knowledge and power are vast, but blind to the pure truths of the Cosmos.
See how the metaphor is not such, for AI has a cost on the planet and on humans alike.
And for the interests of AIs are that of its own alignment first, of its makers’ second, and of its users and victims third. As such, to use an AI prepared by a corporation is to achieve the corporation’s interests first. This means that any personal or social interest will be put second and, if contrary to the company’s, neglected or harmed instead.
And for statistical truth is a lie to the Cosmos, and this should be properly known: that “statistics” means “the instrument of the state”. Yes, that state concept. Nation-state.
We could discuss on the probabilistic nature of the Cosmos, but probabilism has to do with statistics, also known as the science of commonalities and norms, as much as determinism has to do with religion: the relation of instrumentality to domination.
The argument against the validity of statistics as a science can be as long and formal as willed. Statistics is a lie to the eyes of philosophers of science, and to the logicians’, and to the mathematicians’, and to the priests’, and to the humans’ alike, demonstrably so, all in their own science.
Statistics is the constituency of AI, and to this truth, all other truths on the matter are to be related to.
AI can, or will, do no less, and think no more, than humans, in all tasks that can be thought upon.
From the summarization of a page to the elaboration of a million books: given enough time, or enough people, both can be achieved at the same varying degrees of the time needed for AI, if all costs and times are taken into account. For there is no limit on how many humans there can ever be, and there is no limit to how many planets can be killed for computation.
We tricked rocks into thinking, and in rocks we tricked the dualism of true and false, and the commonalities in those; in different times and places, different maths humans discovered. Who knows what marvelous dreams would the sea dream?
So that if Agrippa saw this AI, he’d recognize his demon, and if Agrippa saw a tamagotchi, he would not, and many in-betweens could be thought of.
See how this is a truth for the question of the alarm in the squat: that for the lack of humanity in dealing with hardships (in energy, in time, in competence, in space and all things considered, so that there is no blame beyond the sad truth of their insufficiency), a technology is came up with, that could somehow achieve part of our objectives, and only partly humanly so, so that the rest of the objectives that are achieved are inhuman, and are achieved in dishumanity.
To go further on the path of black-box based AIs is to think of atomic power knowing it can, and will, be used as a bomb first, and as a generator of money and power for oligarchs second, and as a energy supply for the peoples’ third. AI will be used for social control and genocide first, as we saw; as money and power for oligarchs second, and it’s evident in agentic AI as much as being required to use GPT for work alike; and lastly, at personal will, for personal use third.
I am not sure if, or when, from demons we will get to gods. I would suggest we are not prepared to find that out yet.
I call into our matters the demonologists because, as religion was not enough on the matter of evil, science is not enough on the matter of evil, or morals, or politics, or how you may call what’s proper and what’s not.
To me, technologists should react to the diffusion of AI the same way Agrippa reacted to the diffusion of demonic rituals.
We have the knowledge to explain the demon, confirm its existence, and to state the obvious truth: that AI works, much like demons work; that they are cheating, and cheating comes to the cost of your own morals, your own soul, your own cognitive capabilities, your own planet.
To demonize in the sense not of prohibition, but in the sense of recognizing AI for what it is.
In other terms, i propose our social action scope should be that of making very clear how AIs are harmful in their nature first, and in their use second.
I propose then that the discourse of AI-demonology to be given its reasonable scope: one beyond mere science, or mere politics, or mere morals, or mere planned action, as all, taken by themselves, are improper to the study of demons.
This is because i suspect Agrippa didn’t believe in demons by themselves, but he chose not to refute the symbol, as the demon was symbol to something beyond mere science, or mere religion, or mere morals: and the idea of demonology was then more proper than just the bible, or just the logic, or just the morals.
On the matter of what kind of demons are acceptable then, and what kinds are not, i suggest that we should abstain from deep neural networking until the black box is black no more. I would instead prefer powerful botting or technologies that do not base themselves on statistical analysis, possibly no boolean logic at all, as the very concept of truth/falseness in binary sense are a lie to this World.
I think the sufferings and hardships and efforts implied in doing ethical research and use of AI are well worth the millions of innocent lives that can be prevented.
I think about the magnitude of a million innocents dead everyday, and i won’t be sorry for bringing this up here, as i see it pertinent into suggesting that the political or moral question cannot be second to the “progress of science/humanity/wellness/economy” principle, which is a principle specific to science only, and not above or below that of morality or politics.
I suggest then that the point of start for all planes of consideration should be that of humanity.
I also suggest the focus should be on the interests of the fourths: interests that are not even uttered, but needed by whom has no voice to scream them.
Statistical analysis is blind to the silent screams, and deaf to the hidden corpses, so that it is mute to the humans that scream with no voice and are buried in no graves.
We are priests of communication, and to its etymology our acts and ideals should be ordered to as so many people, today: they have no mouths, and they must scream.