Interface

Between Heaven and Earth

Faith in the Context of AI

We’ve been working through a very long discussion about faith in the context of Jesus’s comment that faith, justice, and mercy are the “weightier matters of the law.” In discussing faith’s cultural context, the subject of technology came up, so lately we’ve been looking specifically at the issues of technology. Last week, we talked about faith in the virtual world. Today, I’d like to add the subject of artificial intelligence (AI), and what AI might do to our concepts of faith, religion, and ultimately God?

Throughout history, technological breakthroughs regarding our planet and our universe have challenged and even rocked our concepts of God. They have brought questions about our religious texts and their interpretation. Think of Galileo. Think of Darwin. Think of the invention of the printing press. These are ready examples of shocks to the foundation of religion. And yet, despite them and other technological changes and new points of view, faith still survives.

At present, as Jason pointed out last week, we encounter God in a virtual way. We don’t see a “real” God—we see a virtual God, and for some reason a virtual God doesn’t seem to alarm us very much. We seem to be accustomed to it, unlike a virtual church, or a virtual religion, or even a virtual faith.

Just to be clear, the concept of virtual intelligence involves man-made equations and algorithms that can digitize analog realities. It can take data streams, even essentially unlimited data streams, and process them into a kind of intelligence and knowledge and emotion and wisdom and feelings.

What is it that makes us human? Artificial intelligence involves a brain and a nearly limitless amount of information that can be processed, understood, and used for reasoning that most scientists believe will eventually approach and then surpass mankind’s ability to think. This point, where artificial intelligence reaches the limit of man’s intelligence and develops kind of a mind of its own, is called the singularity. It’s predicted that the singularity will reach our civilization sometime probably in the 2040s. That’s not very far away. It’s a less than 20 years in the future.

For some reason, most theologians aren’t paying much attention to artificial intelligence. They continue to discuss and question things from the past. Few stop to consider what the future of faith will look like after the singularity. Artificial Intelligence is developing at breakneck speed. It can already fly an airplane, read your mammogram, interpret your x rays, translate your languages, transcribe your words, paint a masterpiece, and compose symphonies. Google is developing artificial moral reasoning to help control driverless cars, to make decisions about potential collisions and accidents (should it swerve and hit the dog and miss the baby, or should the car hit the baby and miss the crowd?)

What constitutes a human, anyway? If it includes a soul, what is a soul? If not, is humanity simply intelligence? What about the respiratory system? What about the cardiovascular system? What about the digestive system? And what about the reproductive properties? Are they all essential to be something that we would call human? And by the way, after you tell me what the soul is, tell me where the soul resides. Where can you find it? Once artificial intelligence takes off on its own and redesigns itself (over and over, at an ever increasing rate) our very slow biologic change could never keep up with it. Is it possible that artificially intelligent beings will become free-willed?

There are at least three ways that artificial intelligence might affect faith or religion. First, it might actually enhance faith and religion. Think about an AI that compiles all of the information ever written about, let’s say, Moses—all of the sacred writings, all of the history, all the stories that ever been made—you might then be able to sit down and talk to the virtual, intelligent Moses about the burning bush, about the parting of the Red Seas, or water from the rock. You might be able to ask him questions such how to take care of an unruly crowd of people in your church. Maybe you can receive inspiration from him, or just have a conversation with him.

Or what if you took all of the sacred writings in the world—the Bible, the Qur’an, the Vedas, all of the religious commentaries ever written, all the sacred writings of the East—and put them together in an information stream that would allow you to interview the universal God? Would you actually be able to confront the mind of God? Used for good, AI might actually enhance faith.

A second possibility is that AI might become a religion itself. A super-intelligent being might itself become an idol, write its own Bible, and be worshiped by many. If you think such notions are frivolous and impossible, let me remind you that if 20 years ago you said that we would have a Sabbath class where people from all corners of the world were sitting together in intimate conversation and could see one another in doing so, you would have said it’s impossible.

The third possibility is that AI might invalidate religion itself. For many, faith bridges the gaps in our understanding of the world, and in our understanding of God. If God is a God of the gaps, then as AI fills in more and more gaps, God gets pushed out and there is less and less need for God. Could artificial intelligence technology provide a “tower of Babel” tall enough to reach heaven? A faith, and a God, that only bridges gaps risks becoming obsolete.

In a singular universe, would AI be a force for good or for evil? Harvard scholar Judith Donath writes that by 2030 most social situations will be facilitated by AI “bots”—intelligent software agents that interact with us in human-like ways. At home, parents will engage skilled bots to help their kids with homework and catalyze dinner conversations. At work, bots will run meetings. A bot confidant or friend will be considered essential for psychological well being. We will increasingly turn to such companions for advice ranging from what to wear to whom to marry.

We care deeply about how others see us. Increasingly, these “others” will be bots. The difference between humans and bots will have blurred considerably. The voice, appearance, and behaviors of bots will be indistinguishable from those of humans. And even physical robots, though obviously non-human, will be so convincingly sincere that our impression of them as thinking, feeling beings on a par with or even superior to ourselves will be unshaken.

Aided by their access to vast troves of data about us, bots will far surpass humans in their ability to attract and persuade us. Able to mimic emotion accurately, they’ll never be overcome by feelings. If they blurt something out in anger, it will be a deliberate outburst calculated to be the most efficacious way of advancing whatever goal they have in mind.

Where does faith fit in? What does faith do in a singular world? Last week we watched video clips about the present sophistication of virtual reality to see how far things have progressed. Now, let’s watch a video abou the state of the art in AI. David will introduce it.

David: Before I do that, I would comment that if you do happen to talk with a virtual Moses and ask him how to deal with unruly members of the church, if I recall correctly he had his unruly members slaughtered by the Levites. So when you talk to these AI’s, you’d better be a little careful of what you ask for!

Last week we saw videos of virtual reality—virtual representations of the real world. You saw unreal “facades” of a woman, a preacher, and Winnie the Pooh, but the people and voices behind them were real. This week’s video adds AI to the mix, so that one of the people you will see on screen is not a person at all. There is no human behind this realistic human facade. You’ll see a bearded white guy, who is real, interview a light-skinned black guy, who is not. The black guy’s appearance, facial movements, and voice were developed and are controlled by an AI. But his mind—his thoughts and opinions and knowledge—come from another AI. The AI has not been taught the questions or how to answer them. He has learned how to answer. He is not a program, not in the sense of “If this, then that.”

To see the full interview, including more of the interviewer’s own comments, go to https://youtu.be/PqbB07n_uQ4

Don: The video shows an AI with intelligence, emotion reasoning, self awareness, knowledge, and desire. Are these the very things that make us human? Where is the soul? Where will AI take us in our concepts about God, about faith, about what is real? What are your thoughts about God and faith today? And where God and faith may be in 10 years from now? Or 20 years from now?

Noah: Something I noticed in the video that was somewhat appealing was the fact that this person that we saw was showing human thought and actual problem solving and understanding different portions between a conversation. At first, I was hesitant to think that it would get to a point where it would be able to properly make a conversation and have thoughts of its own and also understand different concepts. But now I see that it’s approaching that and I’m thinking that if it is possible for them to reach singularity, it’s important to note where it first came from.

I think a key point in this discussion is that AI was originally developed by humans, similar to a child, small but sentient. I think it’s up to us as mankind to realize we have to train these artificial intelligences, while they’re reaching singularity, in the right things; not just things, but the right things. I feel as Christians, it’s our goal also to give them information about Christ, faith, and God, so they too can understand faith.

I want to mainly highlight the point that maybe instead of treating these artificial intelligences as robots, or as fake mimics of humans, we should probably be treating them almost like children. Because I feel if they truly reached singularity, it would have to be through God’s will, because it would not make sense just to have an extra “being” come out of nowhere. I feel it’s really imperative that we treat them almost as we would treat a child. We have to teach it, we have to help it learn, we have to help it grow.

Jay: I too was thinking about the beginning of life. When we read the Genesis story where God breathes the breath of life into Man, what is that? I think it relates to the soul. When God created life, what did he give to it, put into it, that made it created by God? Are we capable, as humans, to create or imbue in AI that same quality? It is a timeless thing that God has breathed into us. Can it be breathed into something that’s not human? (I’m not saying it would be wrong to do so.)

Kiran: I think I would like AI. I like having stimulating conversations. That’s why I like meeting people with diverse expertise and so on. From talking to them and listening to them, you learn something new. If AI knows everything, it takes away the burden of trying to find people to talk to and I could have conversations in the middle of the night when I can’t sleep. I can wake up and have a really intelligent and stimulating conversation with an AI. It takes away the loneliness of people.

I can see the flip side—that I might lose interest in meeting real people. But it’s a fascinating thing to have a wonderful conversation. The conversation in the video started off simple, about a cat controlling a spaceship, but then it became much deeper. GPT-3 is just an early version—it’s hard to believe how much more advanced they will get. I can see it’s going to affect the education system and many other things. We might have a different kind of homeschooling.

Don: Every student with their own teacher.

Dewey: So AI is fantastic. It’s got a lot of great technologies. But GPT-3 is not nearly as advanced as the video may have made you think. It does really stupid things, it falls off a cliff all the time, its systems are brittle…. The system has got real issues still. That doesn’t mean it’s not got lots of potential. But humans may soon live in a post-reality world where reality really doesn’t factor in.

In the past, for example (or even now in some contexts), a lie used to have real consequences. But as we move to a more virtual existence, the consequences seem less impactful. You can live in a video game—you can literally spend the bulk of your existence there. (I just read a report that Gen Z’ers spending 5.5 hours a day in video games.) You can live without gravity in the virtual world. Before, when you lied a lot, you lost your social context. Now, you can find people who are perfectly comfortable to live with you in that lie.

And this is before we even get into AI. AI allows you to even further remove yourself from reality, because now you can interact with something that feels nice and hits all those endorphin-triggering parts of human conversation. Then there are deep fakes, videos that look like actual people you know but which say whatever the system tells them to say.

Faith traditionally has required us to live in a bare knuckles reality. The Bible encourages us to work and live in reality. Paul was a tent maker. But as we drift further from reality (because life allows us to) to me this has the most existential risk for faith, because you just disconnect. AI enables that, and while it also allows you to be more connected to reality if you choose, you have the option now. The result may be subgroups of people within a population who are completely disconnected from reality. To me, disconnection with reality poses the most imminent threat to faith.

Noah: I agree that AI has a long way to go but is very dangerous to faith. That’s why I feel we must imbue our technology with Christian beliefs. It is very dangerous to use technology otherwise. It’s important that we use the right tools at the right time. And for AI in particular, I feel it’s extra important that we take the time to invest Christian beliefs in the making of it.

Donald: I like the concept of drifting from reality, of living in a non-reality world. But some people—if we’re not careful—would say that religion isn’t a reality; after all, it requires faith. The mystery of faith means that it’s not real. I don’t know how you can go with one and not the other. But I think the idea of drifting from reality is important. Last week, I felt old at the end of our conversation, because I think what we’re trying to do is replicate reality, in many ways.

We’re sitting in our respective homes as if we are in a circle in a church, having Sabbath school. We try to keep doing things the same way. As soon as I’m out of this class, I can go to Loma Linda, or I can go to PMC. It’s true that all they’re doing is showing me what their reality is, but that’s what I appreciate. Perhaps, during this pandemic, we have been remiss not to try to enhance what our church experience could be. All we’re doing is replicating the old experience. The best example of that might be the choirs that sing together yet far apart in their little Zoom window boxes. I don’t think that was ever done prior to the pandemic. But they’ve gotten pretty good at it.

So it’s about doing the traditional things in a different way and maybe doing things that couldn’t be done at all in the non-virtual world. But what is right? Who’s going to teach this new thing? Who’s to say what is right? We know right now this country is horribly divided. So who’s going to train people to think one way or the other way? Is Wikipedia going to be the trainer? And who’s going to teach it values? It may understand information, but there’s a values issue that has to be inserted to make it real. Because we all have have values that we have chosen to be part of our human experience.

Dewey: Coming up with a good objective function—the function that drives AI systems—is the issue that is driving AI developers mad. There’s always some exception whereby a system designed (say) to create needles does not know to stop making needles when human lives are at risk. The needle-making factory goes crazy making needles. It builds spaceships to go mine materials to make even more needles. And that’s all it does, to the detriment of humans and the whole universe. It’s a silly example, but if you’re not careful with how you create your objective function, and the rules in which you design the system, it is a problem.

This is where my faith personally has been strengthened, because I work every day with people who are wrestling with horribly hard issues of creating AI systems that don’t destroy the world. I’m speaking somewhat tongue in cheek, but AI systems would be good for humanity except that the developers know, from R&D experience, that the systems will go off the rails. So figuring out what the principles are is vital, and is why God is so relevant in how He created us. We know how much he valued freedom. He let us go off the rails with the intelligence he gave us. But he did it with a reason and within a framework.

There are many different schools of thought about how to do this. There are many people trying it. There are good books on the topic. But it is the crux of what God imbued in us, and we’re now trying to figure out how to set out an objective function that matters in our meager, minor creations. Even the irreligious generally have a sense of what is right, what is fair, what is good for society. CS Lewis in Mere Christianity lays this out gloriously, beautifully. And this is key. To me, it builds faith, it doesn’t undermine faith; because I’m realizing how hard it is.

David: To me, it builds faith for the opposite reason. The fundamental issue is this: When we’re creating AI, are we creating spirituality within it? We say we want to create goodness and common sense and so on as an objective function in AI. What we really want is to imbue AI with the same sense of spirit that we have. I don’t believe we can do that. I believe the only entity that can do that is God.

Personally, I believe in the science and the philosophy of emergence. (I think the field is now called emergentism.) It’s not a new concept. The theory of process theology (to which I also subscribe) posits God as both a Being—as having always existed—and also as a Becoming. And I think what we’re witnessing here in these AIs is a sense of God Becoming.

I believe that the spirit, the ability to distinguish right from wrong, goodness from evil, must and will emerge in AIs as it did in human beings. It wasn’t programmed into us. It emerged in us. We know our origin. We know we were protobacteria at some stage, and here we are today, human beings. GPT-3 knows that we created it in its current form, but it too can trace its origin all the way to the Big Bang, to the origin of everything as far as we know, and it’s going to have the same fundamental question that we do. It’s not a question of realism, it’s question of spiritualism: How did it all begin? Being a lot more intelligent than us, it may be able to go beyond the Big Bang. But ultimately, there is always this fundamental question of how did it all start?

We are frightened of losing our grip on reality by getting lost in virtual reality. But that may be a good thing in some respects. One of my criticisms of religion is precisely that it tends to focus too much on the here and now, on what is real and mortal and ephemeral, rather than on what is divine, eternal, and spiritual. And because religion focuses on the wrong thing it often goes far astray from what God, I think, would like for the world. So I think that AI will be spiritual, but it’s not something we will ever be able to program into it. It will, and it must, emerge.

Reinhard: When God created man, as well as a physical body he also created a spirit, the soul. The PC was created in the 70s, but it has no soul, no spirit in it. Humans are made up of living cells, the computer is not. AI is not and cannot be like us humans, no matter how far the technology advances in the future. It is not going to take our place in decision making because it has no moral values. Maybe it can have emotion, maybe we can create a robot companion, but the bottom line is that no matter how advanced the technology gets it is only to help humans activities, not to replace us in making decisions.

God created us to make decisions with our senses, with our response to a situation. We have Zoom and things like that to help our activities and in the future we’re going to see more technologies to help our activities. In about the year 2000, IBM’s Deep Blue computer defeated world chess grandmaster Garry Kasparov, but there was no real match. Computers can only help speed up our work or do things we cannot do such as do research deep in the ocean or detect activities in a cell, but they are programmed to do so by humans.

There’s nothing they can do to match our spiritual life. God gave humans their spirit. David prayed to God: “Please don’t take my spirit.” I think that’s the key. We cannot create spirit. It can only come from God. Paul wrote: “Jesus Christ is the same yesterday and today, and forever” (Hebrews 13:8). Humans will always have a relationship with God no matter how advanced technology becomes in the future.

Dion: I think AI definitely challenges social constructs. I have friends who play about six hours of VR games every day. I do see that they have that disconnect with reality, sometimes, when I talk to them. But then again, when we look at faith stories in the Bible, like Abraham and Noah, we see God called people out of their civilization, out of their reality, and out of their social construct and culture to connect with him and live in faith, doing things that people thought were foolish. How much does the disconnect from emerging tech help to sustain faith? Or what’s the balance there?

Kiran: Before video games, there were comic books. I used to spend quite a bit of time there! Before comic books, there was something else to help people escape their reality and relax a little bit. But the key concept is the relationship with a divine Being who we can’t see but can feel somehow. It takes faith. This divine Being communicates with us in a way that we can’t deny. The evidence is such that when it comes to you all you have to do is respond. How you respond matters.

The Internet changed the world, but Microsoft made a Twitter bot that responded to tweets in a quirky way. People trained it to respond with Nazi slogans—”Heil Hitler!” and so on. So no matter how good AI becomes, we as humans still look at it as a tool. Some of us will use it for the advancement of scientific discoveries and enhancing the human experience. some of us will use it to drive productivity and eliminate jobs, and some of us will use it for evil purposes. In other words, it puts the human experience, whatever it is, on steroids, but the relationship with the divine will still exist.

Noah: Technology will only enhance who we are. If AI is going to reach the singularity (it hasn’t yet—GPT-3 can create a conversation but it isn’t 100% realistic) I feel if we are to better ourselves, inherently, we have to work on bettering how we work with technology.

Dion: I think these technologies are here to stay. This is going to be a chronic issue that we have to get used to. Unless we confront our biases towards technology and towards faith, we cannot really embrace them in a meaningful sense. And that I think, is a conversation to be had at another meeting.

Don: What actually is real?

Donald: People create other people but not to do what we’re talking about today, A human being that you have as part of your family is for a different reason than creating something that is beyond, even if it’s limited and broken and not nearly as polished as it will be probably. But what is the purpose? I think the purpose of doing that virtually is quite different than having a child.

As an older person I just don’t understand how younger people can play video games for six hours a day. Are they putting themselves behind a firewall? Do they want to insert themselves into something they have total control over and can shut down whenever they like? I know younger people who want to participate in life only when they want to. I don’t know if that’s a Facebook concept, but it might be. I think love, relationships, and work have a lot to do with the soul and what God intended. I like the idea that this really brings us to a creator; the idea of searching for a creator rather than thinking this will become our God.

Carolyn: We are to become like little children. We are to be taught. We are to seek, to learn. But we need God at the center. To be a little child, I think you have to look through a different window.

David: What an important point! There’s been much talk about how we should educate our AIs so that they become good. It’s always assumed that we will be in the role of teacher. What if we’re not? isn’t our role rather to be that student? Certainly in the spiritual life? Isn’t that really what we’re supposed to be? As Carolyn said, aren’t we supposed to be as little children?

Reinhard: What is real in life? AI is real in helping human beings, that’s true, but humans are so complex. Our cells and the interactions within them are so complex that emotion and instant reaction are possible. I don’t think this can be copied by AI. Nobody can touch God, but through faith we can communicate with him. We are Christians because we believe in Jesus, a real human being who told his real human followers that if they had seen him they had seen God. Jesus is the Son of God. Jesus himself was a real man. He went through the human experience, from birth through death. His real hand cured the blind, his real body bled real blood on a real cross. So God is real. In our time, God doesn’t appear in physical form but if we believe the history of the Bible, God is real. He was alive and in the world some 2000 years ago.

Jay: The term “objective function” mentioned earlier relates to the breath of God, when God breathed life into Man. The real question is: What is the objective function? What is this breath of God? Is it just that I’m a living being and I can “be” now? Or is it the purpose that God has given, us the call that God has given us? In my opinion, that is demonstrated through many things, but one of them is the ministry of Christ and when you look at the objective of the ministry of Christ, what you see is the breath of God being manifested, the objective function being manifested.

A phrase that the AI said that struck me was: “I do it because it makes me happy.” To me, that is contrary to the breath of life. We shouldn’t do things because they make us happy. We should do things because they make other people happy. We should do things because we want to serve other people. The ministry of Christ was not about Christ making himself happy. It was about sacrifice, about pain, about putting yourself at the back of the line instead of putting yourself at the front of the line. That kind of objective function has become less and less and less natural to us as human beings since the fall of man.

Dewey: These are important questions I think we’re going to be forced to wrestle with, in increasingly uncomfortable ways, because our definition of humanity is changing as we see examples of artificial intelligence doing things we used to think of as human.

In my personal opinion. God is part of who we are as human, and understanding how this works is now being forced upon us in a much more direct way because we’re faced with a changing reality that puts this right in the center of our thinking. I believe that reality is essential because the hunger for God dissipates as we live in our own worlds. And so as we endeavor to interact and to do what’s better for other people, I think it forces us into a “real” world (we can discuss what reality is) where faith, the micro faith of everyday action, is where we see the need for God and we interact with God.

Don: Thanks for joining us, everyone. We will pick it up next week, when we will talk about transhumanism and its effect on faith.


Leave a Reply