Yes, AI and religion can intersect, exploring ethics, morality, and spirituality, while respecting diverse beliefs and traditions.

Apparently, no one is going to swear on the Bible, the Koran and the Torah. That they believe in the superiority of artificial intelligence. But if you find one: it will be really interesting, in terms of worldview.

Due to the ambiguity in this topic: it would be more convenient for me to use either the old name "artificial neural networks" or the name "artificial insects" instead of the frightening name "artificial intelligence". After all, we are talking about the maximum of the same number of artificial neurons as in insects. Yes, they speak. And bees have a language. They draw: and something in a caterpillar allows you to create a complex pattern on a butterfly's wing. Ants and bees: they are also engaged in architecture.

Artificial insects can be dangerous, like some natural ones. I wonder: "locusts with iron faces" from the Apocalypse: is this not a prediction of drones with a neural network? If earlier an official could feel ashamed: now in his place there will be a robot who does not suffer from anything like that. But mostly these are toys.

There is one more question. However, related to the previous one. Since an artificial neural network with such a number of neurons and synapses as a human being absolutely cannot be recorded in the genome or learned in a non-astronomical number of lives to the limit of learning ability - it is simply too big - then either we do not understand how it works at all. Or we have to admit that there is something that stores the necessary information. Let's call it a soul, according to tradition.

The question is that in no country is it taught in school. That a minimum of one and a half million human genomes are needed to record the brain, and so on. This can be taught in an arithmetic lesson and in a biology lesson. The simplest assessments. Question: why is it not taught?

Let's imagine a market for souls. They do not teach in order to bring down the price of souls?

While artificial neural networks have made impressive progress, they are still far from replicating the complexity and adaptability of the human brain. AI systems excel in narrow tasks but lack the general intelligence, creativity, and intuition that define human cognition. Claims of AI surpassing humans often overlook these fundamental differences and should be approached with caution and humility.

I have been involved with this topic since 1986, when we published the back propagation algorithm independently of Rummelhard and Hinton, which did not have much influence outside the USSR, although it did influence early work in the USSR on this topic. And indeed, in recent years there has been dramatic progress, in my opinion primarily in drawing. The fact that neural networks pass the Turing test for text generation is less impressive, since the first programs, not yet neural networks, that passed this test were 60 years ago. Nevertheless, it remains completely unknown how such a large neural network as the human brain works. I believe that we have a dilemma: either we do not understand how it works at all, or there is some distant connection with artificial neural networks, but there is an additional data carrier about this natural neural network, much larger in capacity than the genome, and existing much longer than one human life. In the second case, we can recall some ancient legends.

The fact that the structure of brain of a human and other higher animals cannot be recorded in the genome is a simple and understandable fact that deserves wide fame. Approximately on the same level as data on the planets of the solar system and continents. But there are other, perhaps more specialized, facts. First of all: in 1963, Pontryagin's group was awarded the prize named after the main atheist (Lenin's price) precisely for applying the same mathematics as in neural networks to optimal control. And the corresponding duality theory (it was also known in the West) is mathematically isomorphic to the Hamiltonian formalism of fundamental physics. In physics, the principle of causality is accepted: that is, it is supposed to solve the Cauchy problem and not a mixed boundary value problem, as in optimal control. However. This is an additional principle in relation to the observed facts. In addition, for the stationary Schrödinger equation or, for example, coherent optics of transparent media: the conditions in time are periodic, and in these cases we can directly use the corresponding objects of the material world as adaptive systems for pattern recognition and other problems of this kind.

This is a publication from the time when I somehow did not think about the worldview side of the matter:

https://www.researchgate.net/publication/350801270_1991_Variation_principle_and_algorithm_of_dual_functioning_examples_and_applications_Neurocomputers_and_attention_II_445-452

And it is quite similar we can arrange with any stationary Schrödinger equation.

I am not sure that this especially needs to be brought to engineering implementation. The very fact that in principle any completely isolated part of the physical world, according to modern concepts: in principle a ready-made basis for a neurocomputer - in my opinion is more interesting than engineering implementations.

The feeling is that we see traces of the work of "previous specialists" in artificial intelligence. More precisely, that this is all that we see around. A solution to the Fermi paradox (according to Fermi, we should see traces of the work of highly developed civilizations): we simply do not qualify these as traces due to habit. In ancient legends: by the way, they did qualify. “It’s unknown who discovered the water, but it’s definitely not fishes.”

The topic of manifestations of God's Will is quite complex. For example. As soon as no one swore on the Bible, the Koran and the Torah that they believed that artificial intelligence would surpass natural intelligence: immediately a cheap Chinese chat bot appeared, threatening multi-billion dollar scams in the USA. Is this a coincidence, or did God decide so?

In general, artificial intelligence was conceptually invented by gays. And by the way, it was implemented about sixty years ago. Even before neural networks. It was Alan Turing. He was serving time in a British prison for homosexuality. And he came up with the idea that if an average person in a dialogue with a computer program thinks that he is talking to a person: then by definition, artificial intelligence was created. Then, as tolerance for sexual deviations developed, this led to chat bots. And Trump is heterosexual. But he probably doesn’t know about Turing. Gays could have set him up. One way or another: the topic lacks drive. Let’s do it this way. Maoists train their chat bot. And Trumpists train theirs. Then they will come together in a public discussion on whose country is greater. And whatever they decide, so it will be. You can take bets: who will win.

The hypothesis of the presence of an immortal soul in artificial neural networks, mysticism and world religions: similarities and differences.

The need to introduce an "additional file" storing information when trying to consider the human brain as even a remote analogue of an artificial neural network: arises simply due to the insufficient capacity of the genome (3.3 gigabytes), while an estimate of the file size required to record the structure of the brain (0.63 quadrillion synapses) when encoding communication with 8 bytes requires 5040 terabytes, that is, 1.53 million genomes.

The version that the brain "just learns after birth", if we assume that the brain is not redundant: is also completely untenable. Because the larger the neural network, the longer it takes to train it before reaching the learning limit, which in the case of the human brain leads to an estimate that it is necessary to go through a cycle of all tasks in 15 millionths of a second, and then the next cycle. And so on for 25 years in a row. With such intensity of cramming, the brain would instantly explode like a nuclear bomb. Nothing in common with the carefree childhood of a person.

However: the function of such a hypothetical "additional file" is largely utilitarian. To ensure the formation of the brain and, possibly, the rest of the organism at the same time. By the way: to a lesser extent, but similar assessments are also applicable to higher animals.

On the one hand: religions do not deny that in the absence of a soul, the organism does not function normally. But on the other: in religions, the soul is something far from a utilitarian routine.

In addition, let's say Christianity: recognizes the existence of an immortal soul, but not multiple reincarnations. The Resurrection of Christ itself: can be interpreted as a demonstration of the indestructibility of the essence of man by earthly means.

It seems that early Christians: believed in multiple births. This disappears when Christianity became the state religion. Which gives rise to the hypothesis that it was socially motivated: it is difficult to manage a person who is not afraid of death.

Let us also recall that the duality algorithm from optimization theory, which forms the basis of artificial neural networks: is mathematically isomorphic to the Lagrange-Hamilton formalism of fundamental physics. The only explanation for this today is the artificial nature of the physical world. This does not contradict religions.

It is interesting that today, commanders do not try to explain to the population, in high school, that the structure of the brain cannot be written into the genome: and today's children have very good experience working with memory capacity. They would understand this.

Why we don't remember past lives, and even early childhood.

Despite the extremely superficial similarity of artificial neural networks to the human brain: even such old questions can be looked at in a slightly new way.

The neural network also does not remember the database on which it learned. And it should not. It is: a carrier of action skills. And not a carrier of information about the tasks on which it learned.

By analogy: perhaps a person should not remember previous lives. Or these are separate files that can sometimes be accessed, and this does not have much pragmatic sense.

Then assumptions about past lives can rather be formed by observing a person's skills. By the way, mystics have such a point of view.

This leads to a modification of views on some people. For example. We are accustomed to have a bad attitude towards the Gestapo. In the USSR, the Gestapo is an element of the corresponding folklore.

But there are people. It seems that they have nothing to do with the Gestapo. But if you look: well, a pure Gestapo. By behaviour skills.

Are AI sermons ethical? Clergy consider where to draw the line.

Much of the debate involves grappling with the question of whether AI is being used as a replacement for a sacred human project or whether it's a tool in the service of that project...

“Why not, why can’t, and why wouldn’t the Holy Spirit work through AI?” Lutheran Pastor Naomi Sease Carriker asks. For others, it can circumvent an important intellectual process. “To me, the drudgery is part of the point. I do not want pastors preaching sermons out of Scripture who themselves do not read or study Scripture,” theology professor Brad East says...

https://religionnews.com/2025/07/21/are-ai-sermons-ethical-clergy-consider-where-to-draw-the-line/

The chatbots claiming to be Jesus: spreading gospel or heresy?

Jesus chatbots aren't the only AI technologies seeping into religious practice. Some worshippers don’t agree with the use of them...

https://www.nature.com/articles/d41586-025-02987-9

I'm not a pastor, but I've been interested in the fundamentals of religion. If you already have pastors who don't believe in the immortal soul or in the kinship of man with God, it seems something needs to be done. I've actually noticed something similar about your pastors. And one Orthodox priest complained to me that only a third of the priests believe (he himself was a believer, and incidentally, his first education was in mathematics and programming).

Or do your robots also have an immortal soul? Indestructible by earthly means? Well, it could be a file with data on the parameters of the trained network. But it's easily destructible. For example, on a flash drive. If you hit it with a hammer.

By the way, the key points of this topic can be discussed without mentioning religion. This opportunity is here: https://www.researchgate.net/post/Darwinism_the_Watson-Crick_hypothesis_and_mathematics_are_they_compatible

Darwinism and molecular genetics can be buried. They are dead. And yet, some of you loved them. Darwinism even fed some. You must arrange a decent funeral. Or do you believe in the second coming of Darwin?

Incidentally, this has been discussed with advanced neural network experts. With experience similar to mine, that is, since the very beginning of the neural network boom in 1986. No objections.

As for conference discussions: so far there has only been one Zoom seminar. Also, no objections.

For reasons beyond my control: it was purely by chance that the seminar was held in Jerusalem, the "city of three world religions."

So, personally, I'm betting on the existence of an immortal soul. It's the main conclusion of the experience working with artificial neural networks. You can try to dispute this. Within the framework of science.

I tried to attract those willing to defend Darwinism and molecular genetics through the American presidential social network: https://truthsocial.com/group/darvinism-as-a-satan-project/posts/115214523324448210. Although they are supposedly religious conservatives, and as is well known, there are denominations in the US that do not recognize Darwinism, the result is practically zero. And no one cares, what will replace the lost Darwinism (and molecular genetics).

All this fits well with the content of a forty-year-old text https://www.researchgate.net/publication/350664941_Aliens_Review. The prognoses of this text also match the wars.