Does AI Need a Soul?

Added: 2026-05-27

[00:00:00]Here's a guy who thinks that conscious beings only have value if God says they do.

[00:00:04]>> It's a question you can bring up in just about any conversation you're in. It's a question that helps believers when you're having doubts or you're wondering what if I'm wrong or when God just feels very distant. Now, I'm going to say what that question is at the end of this video. We'll build to get there. I want to start with a fascinating point that Ross Dowet made to Steven Pinker in a recent debate they had.

[00:00:24]>> If AI is able to achieve consciousness, Mr. Dowit, what then? didn't >> well f first of all I'm curious if Dr. Pinker thinks that AI is already conscious since it if if you think that the intelligence of existing AI agents is an indication that humans are not so impressive after all or you know is it be are AI's conscious right now? It >> beats the heck out of me.

[00:00:46]>> No, AI is not conscious. If you understand how AI works and processes information, you can tell the difference between interacting with an AI and interacting with a person. So, it doesn't pass the touring test. Also, we understand the difference between how our minds use language and how AI uses language. When we use words, we index those words to experiences and concepts.

[00:01:05]When AI uses words, it just indexes them to other words. It calculates statistical relationships between words.

[00:01:11]That's not what human minds do. An AI no more understands the words it uses than a dictionary understands the words it defines. I'll link to that debate below.

[00:01:21]Uh, check it out. Fascinating debate.

[00:01:23]But just remember that answer as we go.

[00:01:24]Now, we're going to keep playing the tape. Ask yourself, is what Pinker says next consistent with what he just said there?

[00:01:30]>> Okay.

[00:01:30]>> So, should we assign So, should we we should be agnostic and not kill them?

[00:01:34]Right?

[00:01:35]>> How does that follow? When I care about the well-being of another entity, it's not merely because it's conscious. It's because it's capable of suffering. I can't have empathy for the suffering of another being if there's no possible way that it experiences suffering. Even if AI is conscious, it has nothing analogous to pain receptors or limbic system that could allow it to suffer.

[00:01:53]So, I don't see how there's anything to empathize with.

[00:01:56]>> Well, >> shouldn't turn it off just to be safe.

[00:01:59]>> If there was, I I would not attribute consciousness to the current generation of AIS. I'd feel no compunctions about pulling the plug.

[00:02:07]>> Uh, who knows what would happen if we had a uh, you know, a lieutenant commander data, something that was so convincingly human that it would engage our empathy and make it very difficult to deny consciousness to it.

[00:02:21]>> There's a limit to how much I could empathize with data. if he were real because he is also incapable of suffering unless his emotion chip is activated. However, I think data is a bit unrealistic because he doesn't have emotions but he expresses desires and I'm not sure you can have desires without emotions.

[00:02:36]>> Personally, I don't think it would be conscious. Why not?

[00:02:38]>> But I couldn't. Uh the the thing about consciousness in the sense of pure sensience, first person subjective experience, is that by definition you can't tell if anyone else has it, including of course other humans.

[00:02:55]>> No, that's not true by definition. By definition, you can't experience other people's experiences, but you can reasonably infer them from observing phenomena which consistently correlate with certain experiences. That is, it's impossible to tell if any other human is conscious.

[00:03:10]>> Aren't you sawing? Aren't you sawing off now the branch on which your entire philosophy rests? You've spent the whole evening telling us it's self-evident we don't need Christianity. It's self-evident that we should care about the well-being and feelings and experiences of other human beings. And now you're telling me that you can't even tell if I have those feelings at all.

[00:03:28]>> I don't agree that it's self-evident that you should care about other people's feelings. I do care, but I don't think that there's any objective reason why I should. I care because my brain is wired in such a way and my experiences have programmed it in such a way that I can't help but care. That's why I don't need Christianity as a moral framework. In fact, I don't see why any objective moral framework would matter.

[00:03:48]Because if that framework isn't consistent with what my conscience makes me care about, then I won't be capable of caring about it. Even if there is a God that says I should care about something, that's not necessarily going to cause me to care about it. Well, >> this is what I mean by the self undermining nature of the materialist project, right? The worldview doesn't fit together. You have pieces that don't fit.

[00:04:09]>> So, remember that language from Ross, pieces that don't fit. Perhaps Steven Pinker is not the best representative of physicalist conceptions of morality because I don't think his arguments are very good either. What doesn't fit about my view though? Also, I don't see how theism solves this issue since even if there is a God who dictates morality to us, there is no objective non-ircular reason to do what God says.

[00:04:31]>> Okay, keep that in the back of your mind. We'll come back to that. I think Ross is tapping into an inconsistency here or a sense of disjointedness in an atheistic worldview or at least most expressions of atheism. Put it like this. If you're not sure when AIs become conscious, why not heir on the side of caution and never terminate them unless you had to or something like that?

[00:04:54]>> Because I don't care about consciousness per se. I care about suffering. I empathize with entities that have desires and not just awareness.

[00:05:02]>> Once that question came into view, Pinker seems to pull back a little bit in his answer and he's saying, you know, personally, I don't think a lieutenant commander data would be conscious. And in the context of their discussion, he seems to be appealing to an intuition most of us have that human beings are special.

[00:05:18]>> If an android can elicit empathy, then clearly this empathy is not an intuition that humans are special. In fact, there are plenty of nonhuman animals that elicit empathy and even produce the intuition that they're conscious. So, how is that an intuition that there's something special about human beings?

[00:05:34]>> There's something valuable about human beings.

[00:05:37]>> I find at least some value in preventing the suffering of anything that can suffer. Not only does that extend beyond humans, it doesn't even extend to all humans. Zygot, embryos, and even a lot of fetuses show no evidence of being capable of suffering. So, I don't value them as much as I value organisms that do.

[00:05:53]>> But in the context of a strictly naturalistic view of atheism, like what Pinker espouses, it's unclear why that would be.

[00:06:01]>> Humans have value because I value them.

[00:06:03]I don't see why I need a god to exist in order to do that. Suppose an AI did become just as intelligent as us. You're not sure if it's conscious, but it betrays all the features of consciousness. Why wouldn't it merit just the same ethical obligations that human beings have? It's a fascinating question because consciousness alone doesn't make doing so worthwhile. Again, lots of non-human animals seem very conscious, and we grant them at least some consideration, but we don't give them all of the same rights as humans.

[00:06:32]Because while animals look like they can suffer, humans can have things like hopes and ambitions that animals, as best we can tell, are not capable of having. That adds an extra kind of suffering that looks like it's unique to humans and maybe a handful of additional species.

[00:06:45]>> Now, earlier in their discussion, Pink are grounded our ethical obligations in something like the golden rule.

[00:06:51]>> We all seek life, health, and happiness.

[00:06:54]>> No, we don't. Most people probably do, but not everybody. And even if we did, things that make different people happy can not only be different, but sometimes even mutually exclusive.

[00:07:03]>> And if we claim them for ourselves, we cannot deny them to others.

[00:07:07]>> How does that follow? The only reason I don't deny them to others is because it's unconscionable to me. But it's not unconscionable to everybody. And I see no objective reason to claim that this is wrong. I don't like it because I don't like hypocrisy. And I would vote for laws that force consistency in this regard, but I don't claim any objective justification for doing so. We have a capacity for reason which can be amplified by education, science, and free inquiry. This gives us a common mission to apply science and knowledge to reduce suffering and enhance flourishing. Why though? I like all of those things, but I see no objective reason I ought to. This is the case even if there is a God who says that this is our mission. I see no objective non-ircular reason why we have to do what God says.

[00:07:51]We know this can make us better off because it has made us better off. We have doubled our lifespans, decimated poverty, and slashed the toll from crime, war, and accidents. No thanks to God, nor to an indifferent universe, but rather to human reason applied to human flourishing.

[00:08:10]Thank you. Let me put his opening words up on screen here. We all seek life, health, and happiness. Here's the logic.

[00:08:16]And if we claim them for ourselves, we cannot deny them to others. So this is an appeal to the golden rule essentially. But the golden rule is a tool for determining what moral behavior should look like. It doesn't in itself explain why moral obligation exists in the first place. Nor does God. Why are we morally obligated to do what God says? Where does that moral obligation come from? If it comes from God, then the argument is circular. Are we morally obligated to care about other humans because they're created in the image of God? Why does that matter? Is it because God says it matters? Theistic morality is entirely circular.

[00:08:54]>> It's more of a moral guide than a moral foundation. The big question is why should we treat our neighbor as oursel?

[00:09:02]>> For you. The big question should be why should we do what God says?

[00:09:06]>> Why are you obligated to care about others life and health and happiness just because you care about your own? I don't believe there's any objective obligation. I care because my conscience cares about consistency most of the time. However, that's just an explanation of my morality. It's not a justification. I don't believe that there are any ultimate justifications for morality. The step from I care about these things to therefore I'm obligated to treat others with sensitivity to those things is not self-explanatory.

[00:09:37]>> That's right. It doesn't explain how to get an ought from an is. However, it's also not self-explanatory why you ought to value a human just because that human is created in the image of God. You can't get an ought from that is either.

[00:09:48]>> That's certainly not how the animal kingdom works. You know, if you have a a chimpanzee community carrying out a violent raid against another chimpanzeee community, you would need to say something more. Say there's a a chimpanzeee community on the north north side of the river going over and slaughtering all the the chimps on the south side of the river. Um, if and you wanted to say, "Well, you ought not to do that." You'd need to say something more than just the chimps on the south side of the river don't like that.

[00:10:14]>> You'd also have to say something more than God doesn't like it.

[00:10:18]>> They don't want to be slaughtered. Well, of course they don't. Uh, that doesn't establish genuine moral obligation in and of itself. When we look at human civilization, we have the same need. You need some kind of foundation for moral obligation that goes beyond this generally helps society. and something that can account for our particular obligation to human beings. Again, an advanced AI might also seek life, health, and happiness. If I were convinced that such an AI exists and is capable of suffering and the kind of suffering that is typical of humans, then I would care about it as much as I would care about any human, would a Christian not care? Would they lack empathy for this AI person just because it isn't human? Now, a theist has a metaphysical ground that explains in a satisfying way moral obligation. Here's how Ross puts it.

[00:11:08]>> It's good to believe that you should love your neighbor as yourself because your neighbor is made in the image of the creator of the cosmos.

[00:11:14]>> How does that follow? That seems like a complete nonsequittor to me. Even if you believe in God and you value God, why does it make sense to value everything that resembles God? Why does that matter? Gavin said that this would be a satisfying explanation of why we should care about other people. But so far, I'm not satisfied. The whole argument is effectively, "God says so. Why should we treat each other equally? God says so.

[00:11:35]Why should we respect each other's rights? God says so." A question they never answer is, "What non-ircular reason is there to believe that we ought to do what God says?" And it's good for human beings to believe that human life itself has a special and particular value, particularly in a moment when digital technology and artificial intelligence threaten to make many human lives feel almost obsolete.

[00:12:00]>> That makes it sound as though if an AI existed that could have desires and could suffer, this guy thinks it would be a good thing to disregard that suffering. My conscience could never get on board with that. So this is the old idea of creation in the image of God. We just take this for granted today. I find in the modern west. This idea is implicit in so much of our moral reasoning. But sometimes we don't even notice it. So take for example the second sentence of the Declaration of Independence. And this amazing statement that we're created equal and endowed by our creator with certain rights. Let's say God created a planet with people to whom he gave rights, but he doesn't step in to enforce them. So, some people believe that they have God-given rights, but others don't. And let's say he created another planet with people to whom he did not give any rights, but nonetheless, some people believe that they have God-given rights, and some don't. How would you be able to tell which is which? Of what consequence are these God-given rights if he doesn't enforce them? How is their existence discernible from their non-existence?

[00:13:11]To everyone who helps me out on Patreon, you're a big help. Thanks so much.