I’ve been doing a lot of thinking about AI. It’s hard work. AI seems to have many tenacles, many rabbit holes down which one can fall. Many of those holes seem to be filled with foxes instead of rabbits. Nothing cute and cuddly down there.
It has also been worsened by media and hype. Those who are behind AI are companies, and companies exist to make money for their owners, shareholders, and the like. In our world, hype means exposure means more money. So, it is in the interest of AI producing companies to keep up the hype.
That helps precisely no one. It is just greed.
Confession. I am a late adopter by personality. I don’t like the rapid change of technology, I find that adopting new tech updates too early leads to confusion and frustration. I have never experienced anything that has suggested I reconsider this bent of personality that I am increasingly careful to maintain. Nothing. Call my cynical, but I don’t think that’ll change.
Nevertheless, in my thinking on AI, I have found some help in a podcast discussion on Exodus 7. In Exodus 7, Moses and Aaron go to meet Pharoah to tell him that God commands him to let the people of Israel go. Pharoah will demand a miracle, says the Lord, and when he does, Aaron is to throw his staff down and it will become a snake. This is just what happens.
But Pharoah summons his magicians, and they preform a trick with their staffs and they become snakes when thrown down. Aaron’s snake eats the magicians snakes, and Pharoah hardens his heart.
Then, the Lord tells Moses to go to the Nile, strike it, and it will become blood. Moses does so. Pharoah calls his magicians, and they appear to do the same with their secret arts, so Pharoah hardens his heart, comforted that his magicians can turn water into blood (or appear to).
The point made on Bible Talk podcast I was listening to is that Pharoah’s magicians were basically useless to Pharoah. On first look, it might seem like they have the same power as Moses and Aaron, but when we stop to think deeper, what is it that Pharoah needs? He needs his magicians to do something original, not to mimic what God has done through Moses & Aaron.
In both scenarios, the magicians can only mimic God’s power. They have no originality, no power of their own, nothing creative about them and their supposed power. This is crystal clear in the Nile episode: what Egypt really needed from their magicians was to turn the bloody Nile back into water. This they could not do. They could only mimic what God did through his servants Moses and Aaron.
All evil is destructive mimicry. Satan has no creative power of his own. Only God can create. Satan sees and mimics. Badly. Satan and his minions (which sadly include humans who willingly pursue evil) can only mimic and it’s never helpful or useful. But it is so often deceptive.
Yet it is deceptive only to those who have a reason or disposition to be deceived. So Pharoah. He does not want to let his slaves go free – he has a vested interest in keeping their labour. So he wants to see his “gods” do miracles. He’s ready to believe in his magicians magic.
But magic is not real. The story in Exodus makes it clear that Pharoah and his magicians are incompetent. Their best magic is useless (their snakes are eaten by Aaron’s) and unhelpful (they make more ‘blood’ that is undrinkable) and only increases the suffering of the Egyptians, thereby demonstrating God’s glory and power.
What’s all this got to do with AI?
Well, I think the same thing is happening with AI when it comes to discussions around sentience and consciousness. Quite apart from the need to define those terms, and the philosophical and worldview battles that are going on around what it means to be human when we use such terms, there is also the issue of magic and mimicry.
Edward Freser makes the point (here and here) that the arguments about AI and intelligence/consciousness are akin to magic. Here’s the key passage (long, but worth it):
Here’s an analogy (which I have developed in more detail elsewhere). The methods employed by entertainers such as David Copperfield, David Blaine, and Penn and Teller are designed to produce effects that merely simulate magic. And no matter how well they work, we know that mere simulation is all they can achieve, because the means they use are in no way preternatural but entirely mundane – sleight of hand, illusions, and so on. Of course, genuine magic is not real in the first place, but that is irrelevant. The point is that even if it were real, it would not be what Copperfield, Blaine, and Penn and Teller are doing, precisely because their methods are not of the type that could produce more than mimicry.
Now, AI operates on an analogous principle. It is designed to produce effects that simulate intelligence, by means that don’t require any actual intelligence on the part of the machines themselves. And this is as true of machine learning and related approaches that now dominate AI research as it was of the Turing machine model that dominated the earlier history of AI. To borrow some terminology from John Searle, AI algorithms of whatever kind are sensitive only to the “syntax” of the representations they process rather than their “semantics.” That is to say, they process representations in a way that is sensitive only to their physical properties rather than the meanings that we associate with those physical properties.
Because there is a general correlation between syntax and semantics – for example, the word “love” on a printed page is typically going to be used to express the concept LOVE – the output of a sophisticated algorithm can be made to simulate the sort of thing a human being might write. The correlation is not perfect. For instance, there might be cases where the string of letters “love” is not actually being used to express the concept LOVE, and there might be cases where the concept LOVE is being conveyed but without using the string of letters “love.” This is why AI programs can often be “tripped up” and reveal, through a failure to reflect such nuances, that they don’t actually understand what they are “saying.” Exposure to further data or refinements to an algorithm might work around such problems. But all that that yields is a more convincing simulation (just as Penn and Teller and company may come up with new ways of producing ever more convincing illusions). It doesn’t somehow generate a grasp of meaning or semantics on the part of the machine, because its basic structure is in no way sensitive to that in the first place.
Our authors speak as if the dispute over whether AI models are genuinely intelligent or merely simulating intelligence is a matter of “inference to the best explanation” – as if their position and that of the AI skeptic are alternative hypotheses meant to account for the same evidence, and the controversy has to do with which view is more empirically adequate, more parsimonious, and so forth.
But this is as silly as suggesting that the view that Penn and Teller are merely simulating magic rather than producing the real thing is being proposed as the “best explanation” of the “evidence.” It is, of course, nothing of the kind. It is just a simple and straightforward conceptual point about the nature of their methods. And to note that AI produces only a simulation of intelligence rather than the real McCoy is an equally simple and straightforward conceptual point about the nature of its methods.
AI is only able to simulate intelligence.
AI is only able to mimic consciousness. And this will always be true. Such mimicry might become ever more sophisticated, and as such might be able to deceive many more people, but it will remain only mimicry.
The dangers in such convincing mimicry are, I hope, obvious. We see it in the rise of deep fakes. We see it in the rise of AI-girl/boy friends. But the danger is not just in these outworkings, it is in the fact that AI encourages us to desire being deceived in such ways. Much like magic, we enjoy the ease and entertainment of such deception – for AI can give us what we want when we want it, with no moral qualms about whether we should want it in the first place!
Without such boundaries, we get used to being deceived, and like anything else addictive, we have withdrawal symptoms when we encounter friction in real life.
So, relationships. An AI friend is always available, always says the right thing, always agrees with you, always comforts you. There’s no friction, no boundaries, no one to question your motives, desires, opinions, decisions. Nothing but a magic mirror telling you what you want to hear. And that leads some to make utterly tragic and desperate choices, like taking their own lives.
This is Satanic. We must not be fooled. Whilst AI as such might not be satanic, it is a technological tool bent toward harm, which satan will do all he can to use for his lethal, destructive purposes. We should not fear this – for Jesus has defeated the work of satan in his death and resurrection on the cross. But we should pay attention to what Jesus says about Satan, and not assume that AI is neutral.
He was a murderer from the beginning, not holding to the truth, for there is no truth in him. When he lies, he speaks his native language, for he is a liar and the father of life.
Photo by Mick Haupt on Unsplash
Holding on to Hope 