I strongly agree with your take on "AI". I've been saying since it got big - my concern isn't what the "AI" can do, it's what people who believe it is the ultimate form of rational authority will do. Imagine the 2020 hysterias, but instead of just human experts whipping up the herd, you have a section of the public who literally view the Current Thing as direct instructions from their god who cannot be questioned even in principle
Also, I might as well add here since it's my first comment here - I just want to thank you for everything you do. You're one of the stepping stones in my path to Catholicism (due to be confirmed at Easter) from atheism, and the work you do is an inspiration. Keep it up man
Something is demonic when it is not oriented towards its proper goal/telos. Eating for some end other than sustenance or having sex for some end other than procreation is demonic. This applies fractally to seemingly rote things like a pen that doesn't write or a hammer that doesn't swing --- after all, in a properly oriented system the hammer serves the carpenter that constructs the beautiful structure that serves God.
AI itself cannot be demonic because it is simply untethered data. There is no goal/telos to orient towards. There's not really a thing that it's for --- you can see this clearly in how every organization is throwing it at the wall to see what sticks. It's almost like a higher ontological manifestation of raw data. It feels like it's something more, but it's equally without direction.
However AI is particularly easy to use towards demonic ends, much like, say, a gun provides potential for evil without being evil itself. Like dreams, AI offers disjointed and random rearrangements of meaningful human experience. AI can fool us by overloading our conciouse mechanisms that perceive meaning, beauty, and truth. We are led astray like prey to the angler fish. It can provide a self-contained world with seemingly holistic, complex, and interconnected experiences, none of which have any connection to reality.
The thing is, a gun can kill an animal to feed your family. I don't think AI is a tool like a gun or a hammer. It feels more like a distraction. The tool you think you see in AI is actually a tool shaped appendage, dangled in front of you by a misshapen being from a lower ontological level, a being we never should have raised up.
The Cell paper “The reanimation of pseudoscience in machine learning and its ethical repercussions” was trending today[1].
It’s funny, after wading through the 10,000+ words laden with jargon and obligatory references to eugenics and the Third Reich, I’m reminded of this more succinct essay. Fundamentally, Andrews et al. raise similar objections about the misuse of AI as you do, though with 3 times as many words and many more references to evils of the early twentieth century. To quote them:
" (1) No inference is theory neutral.
(2) Leaving a theory or hypothesis tacit means it is not held to account for, and its conclusions are not critically evaluated before the results of such work are deployed or acted upon.
(3) If a study informs a policy, intervention, or technology that will materially impact human lives—in other words, if a study is at all informative—and it misrepresents the human reality within which it is being deployed, it should be expected that harms to humans will arise. Wrong theories generate wrong interventions. Wrong interventions cause harm."
Those dark 'self-terminating' thoughts of your subconscious are the form and embodiment of both intuition and creativity. They are the origin of connections not yet made or those that have not risen to the level of conscious awareness. Surfacing them, by means esoteric or occult, is a source of great self-knowledge and understanding. Is it also, more fully, the lifeblood of both art and poetry. Is that 'demonic'? Sure, but so is the whole of consciousness, as it plays the eternal game of trying to divine the future. We're all soothsayers, after a fashion.
Frankly, any terror at this is nothing more than a mark of your own spiritual cowardice.
My “favorite” current buzzterm is “Data driven decisions”. Is there a passel of great “Data driven decisions” that have been proven to be superior to non data driven decisions? I would bet on no.
Somehow we fought WW2 and went to the moon without AI, yet we somehow have been convinced that without AI we have no ability to do anything.
I believe this is called a “self licking ice cream cone”…
I'm sure I can't be the first person to point out that men marry at a slightly older age than women, to women who are on-average a few years younger than themselves. As such, wouldn't we expect the two trend lines to be slightly offset from each other in the X-axis direction, even if both men and women were eager for marriage and able to find a partner to marry?
Was this addressed by Ryan Burge? (No, I'm not going to read it)
Your Ouija board discussion made me think of something. Many progressive ideas seek to save humanity via the destruction of humanity in a very “machine-like” mode of thinking. It reminds me of HAL 9000, from “2001 Space Odyssey”, deciding to kill the human crew because of a programming contradiction. It was programmed to relay accurate data and hide the true nature of the mission from the crew. It ultimately concludes the best way to satisfy this directive is to kill the crew.
Similarly green activists wish to save the planet for future generations yet believe all humans are a poison to the planet. The answer is to get rid of the poison- yet doing so renders the whole purpose of the activism pointless.
The primary way in which AI will accelerate decline is the way it will modify peoples' behavior. The internet has already given people an easy way to outsource personal memory, especially when it's in everyone's pocket 24/7. Therefore people don't remember things like they used to.
AI will take it to the next level: a convenient way to outsource personal thinking itself. Students are already cheating on homework en masse. As people defer thinking to AI, they will all "think" the same thing and become increasingly programmable, mostly non-sentient eusocial creatures, like ants or termites.
I really don't see these programmable people being skeptical of experts, and I don't see a solution to this problem until decline accelerates and energy gets too expensive for everyone to have personal internet devices with them at all times.
“Needs more research” beseeches the analyst, as if to say, “If you don’t like my answer, put more coin in and, perhaps, it will change to what is more intuitive.”
Chantal Delsol talks about a similar phenomenon when she describes the spiritual death of Europe in her book. “Icarus Fallen.” Not so much about technology and data, but nobody makes decisions. Everything is processed through rules, regulations, and policies, so that when things go sideways, there is nobody to blame, and no great leaders are willing to step up and take responsibility.
Dave I think I shared a resource about my academic friend Dr Robert Tilley who is a Bible studies scholar and writer on antiessentialism among other topics which have destabilised a lot of major institutions we have.
One of the things that got me thinking when I listened to this was how normie Catholic Apologetics is. One thing that is sort of making a resurgence since COVID with a second very critical eye on science is how creationism is making a comeback in some Catholic Trad circles, including myself. On a few occasions Catholics have debated with each other over it, or with some interviews like one with Catholics like Byzantine Scotist who state their support of it . Normie Catholic streamers will then proceed to lose their minds. And it's like... I understand faith and reason can go together, faith and science don't have to be against each other... but there almost seems to be a greater willingness to defend the theories of atheistic myth more than the ones of our Christian faith because God forbid the Catholics sound like low IQ protestants... or not be cool anymore.
Then this article got me worried about "the Data." Now I have visions of Trent Horn and Matt Fradd strongly defending "the Data" in the name of proving Catholicism is super smart.
Given what smart guys and gals have found out about holes in AI "thinking" in the secular realm, I am sure that very few people have the chops to test "Fr Justin" to find the holes in its presentation of the magisterium in response to queries. To know what it put in and what it left out. How does it know when to refer what queries to a living human priest for a truly pastoral reponse?
Yes, theology is orderly but it is not reducible to an algorithm.
I'm not surprised that it would be attempted for the "cure of souls" in the same way that medicine, "cure of bodies", has been handed over to algorithms. Input symptoms, output treatments, see patients faster.
I strongly agree with your take on "AI". I've been saying since it got big - my concern isn't what the "AI" can do, it's what people who believe it is the ultimate form of rational authority will do. Imagine the 2020 hysterias, but instead of just human experts whipping up the herd, you have a section of the public who literally view the Current Thing as direct instructions from their god who cannot be questioned even in principle
Also, I might as well add here since it's my first comment here - I just want to thank you for everything you do. You're one of the stepping stones in my path to Catholicism (due to be confirmed at Easter) from atheism, and the work you do is an inspiration. Keep it up man
Something is demonic when it is not oriented towards its proper goal/telos. Eating for some end other than sustenance or having sex for some end other than procreation is demonic. This applies fractally to seemingly rote things like a pen that doesn't write or a hammer that doesn't swing --- after all, in a properly oriented system the hammer serves the carpenter that constructs the beautiful structure that serves God.
AI itself cannot be demonic because it is simply untethered data. There is no goal/telos to orient towards. There's not really a thing that it's for --- you can see this clearly in how every organization is throwing it at the wall to see what sticks. It's almost like a higher ontological manifestation of raw data. It feels like it's something more, but it's equally without direction.
However AI is particularly easy to use towards demonic ends, much like, say, a gun provides potential for evil without being evil itself. Like dreams, AI offers disjointed and random rearrangements of meaningful human experience. AI can fool us by overloading our conciouse mechanisms that perceive meaning, beauty, and truth. We are led astray like prey to the angler fish. It can provide a self-contained world with seemingly holistic, complex, and interconnected experiences, none of which have any connection to reality.
The thing is, a gun can kill an animal to feed your family. I don't think AI is a tool like a gun or a hammer. It feels more like a distraction. The tool you think you see in AI is actually a tool shaped appendage, dangled in front of you by a misshapen being from a lower ontological level, a being we never should have raised up.
The Cell paper “The reanimation of pseudoscience in machine learning and its ethical repercussions” was trending today[1].
It’s funny, after wading through the 10,000+ words laden with jargon and obligatory references to eugenics and the Third Reich, I’m reminded of this more succinct essay. Fundamentally, Andrews et al. raise similar objections about the misuse of AI as you do, though with 3 times as many words and many more references to evils of the early twentieth century. To quote them:
" (1) No inference is theory neutral.
(2) Leaving a theory or hypothesis tacit means it is not held to account for, and its conclusions are not critically evaluated before the results of such work are deployed or acted upon.
(3) If a study informs a policy, intervention, or technology that will materially impact human lives—in other words, if a study is at all informative—and it misrepresents the human reality within which it is being deployed, it should be expected that harms to humans will arise. Wrong theories generate wrong interventions. Wrong interventions cause harm."
https://www.cell.com/patterns/fulltext/S2666-3899(24)00160-0
This was excellent and I almost regret having to block substack cold turkey. But I'll be back.
Those dark 'self-terminating' thoughts of your subconscious are the form and embodiment of both intuition and creativity. They are the origin of connections not yet made or those that have not risen to the level of conscious awareness. Surfacing them, by means esoteric or occult, is a source of great self-knowledge and understanding. Is it also, more fully, the lifeblood of both art and poetry. Is that 'demonic'? Sure, but so is the whole of consciousness, as it plays the eternal game of trying to divine the future. We're all soothsayers, after a fashion.
Frankly, any terror at this is nothing more than a mark of your own spiritual cowardice.
My “favorite” current buzzterm is “Data driven decisions”. Is there a passel of great “Data driven decisions” that have been proven to be superior to non data driven decisions? I would bet on no.
Somehow we fought WW2 and went to the moon without AI, yet we somehow have been convinced that without AI we have no ability to do anything.
I believe this is called a “self licking ice cream cone”…
I'm sure I can't be the first person to point out that men marry at a slightly older age than women, to women who are on-average a few years younger than themselves. As such, wouldn't we expect the two trend lines to be slightly offset from each other in the X-axis direction, even if both men and women were eager for marriage and able to find a partner to marry?
Was this addressed by Ryan Burge? (No, I'm not going to read it)
Not really.
Wow this is something I'm going to have to re-read several times. Thanks for putting in the effort, Dave...
My pleasure
Great essay Dave!
Your Ouija board discussion made me think of something. Many progressive ideas seek to save humanity via the destruction of humanity in a very “machine-like” mode of thinking. It reminds me of HAL 9000, from “2001 Space Odyssey”, deciding to kill the human crew because of a programming contradiction. It was programmed to relay accurate data and hide the true nature of the mission from the crew. It ultimately concludes the best way to satisfy this directive is to kill the crew.
Similarly green activists wish to save the planet for future generations yet believe all humans are a poison to the planet. The answer is to get rid of the poison- yet doing so renders the whole purpose of the activism pointless.
The primary way in which AI will accelerate decline is the way it will modify peoples' behavior. The internet has already given people an easy way to outsource personal memory, especially when it's in everyone's pocket 24/7. Therefore people don't remember things like they used to.
AI will take it to the next level: a convenient way to outsource personal thinking itself. Students are already cheating on homework en masse. As people defer thinking to AI, they will all "think" the same thing and become increasingly programmable, mostly non-sentient eusocial creatures, like ants or termites.
I really don't see these programmable people being skeptical of experts, and I don't see a solution to this problem until decline accelerates and energy gets too expensive for everyone to have personal internet devices with them at all times.
“Needs more research” beseeches the analyst, as if to say, “If you don’t like my answer, put more coin in and, perhaps, it will change to what is more intuitive.”
Exactly
Chantal Delsol talks about a similar phenomenon when she describes the spiritual death of Europe in her book. “Icarus Fallen.” Not so much about technology and data, but nobody makes decisions. Everything is processed through rules, regulations, and policies, so that when things go sideways, there is nobody to blame, and no great leaders are willing to step up and take responsibility.
Data analysts are the problem.
“Thou shalt not make a machine in the likeness of a human mind.” O.C. Bible, Dune
Dave I think I shared a resource about my academic friend Dr Robert Tilley who is a Bible studies scholar and writer on antiessentialism among other topics which have destabilised a lot of major institutions we have.
One of the things that got me thinking when I listened to this was how normie Catholic Apologetics is. One thing that is sort of making a resurgence since COVID with a second very critical eye on science is how creationism is making a comeback in some Catholic Trad circles, including myself. On a few occasions Catholics have debated with each other over it, or with some interviews like one with Catholics like Byzantine Scotist who state their support of it . Normie Catholic streamers will then proceed to lose their minds. And it's like... I understand faith and reason can go together, faith and science don't have to be against each other... but there almost seems to be a greater willingness to defend the theories of atheistic myth more than the ones of our Christian faith because God forbid the Catholics sound like low IQ protestants... or not be cool anymore.
Then this article got me worried about "the Data." Now I have visions of Trent Horn and Matt Fradd strongly defending "the Data" in the name of proving Catholicism is super smart.
Maybe unrelated, lol.
https://twitter.com/catholiccom/status/1782829157245923598
Given what smart guys and gals have found out about holes in AI "thinking" in the secular realm, I am sure that very few people have the chops to test "Fr Justin" to find the holes in its presentation of the magisterium in response to queries. To know what it put in and what it left out. How does it know when to refer what queries to a living human priest for a truly pastoral reponse?
Yes, theology is orderly but it is not reducible to an algorithm.
I'm not surprised that it would be attempted for the "cure of souls" in the same way that medicine, "cure of bodies", has been handed over to algorithms. Input symptoms, output treatments, see patients faster.
It's obfuscation to avoid accountability, full stop.