Can AI ever be conscious?

Added: 2026-05-12

[00:00:00]And I'm um I'm unsure if this is working.

[00:00:35]Hello. Um, I'm not sure if I'm connected.

[00:00:39]Yeah. Okay. So, uh, people tell me I'm coming through. There was some software was acting up on me there. I wasn't sure I was connected. So, welcome. Um, back from my retreat.

[00:00:52]Um, I'm going to be uh continuing this Thursday evening broadcast for uh foreseeable future, but there's going to be um another interruption at towards the end of the month. I'm I'm going to be traveling, but that's okay.

[00:01:10]I'll I'll I'll be back. I think it's only two weeks.

[00:01:14]So, um I didn't want to start a series because I'm going to be interrupting it again fairly soon. So, I'm going to do a oneoff this week and and um uh subsequent broadcasts. Now, I thought of this um talking about AI and consciousness uh first of all because consciousness is uh a very um central concern in Buddhist uh thought and uh and in practice. uh to try and understand the nature of consciousness.

[00:01:55]And there has been some talk about this lately apparently um Richard Dawkins who's a well-known uh uh popular science popularizer. He's written a number of uh quite good books on on evolution. that he's uh by his own personal philosophy is very uh hardcore materialist or physicalist and um he recently said that he spent several hours interacting with an AI system known as quad. I hadn't heard of that one before and uh came away with the um being convinced that it's actually conscious.

[00:02:55]I think this um this comes from uh his own uh personal philosophical bias that he sees consciousness as something arising from from materiality.

[00:03:16]So, uh there's no reason why it couldn't arise in a uh a a software system running on uh running on the computer.

[00:03:35]And part of I think part of the um the problem with the the the debate or the question about uh uh consciousness and uh artificial intelligence is a confusion between consciousness and uh intelligence.

[00:03:55]In um Buddhist thought there's a very clear idea of what uh of a definition for for consciousness uh chitta or wana.

[00:04:13]Uh however even within Buddhism there are different schools of thought different debates about nuances and I may uh I'm thinking maybe next week I'll get into that.

[00:04:26]But um uh consciousness or uh let's say the the chitta is the subjective side of of mentality. It's in Thai they call a puru the one who knows that which has an experience.

[00:04:50]um it isn't a manipulation of data or algorithms. No, that's intelligence.

[00:04:57]And that would be part of in the um Buddhist psychology that would be part of uh the uh aggregate of sankara's mental formations and be classed as a in abidama would be classed as a chitasa.

[00:05:14]Abidama divides the mind into two sides chitta and chitasa.

[00:05:20]Chittasaga just means that which goes along with chitta. Chitta is the one who knows. See the having an experience.

[00:05:37]There's one modern um writer on uh theory of mind and the psychology of consciousness and I um I can't recall his name at the moment but he came up with uh uh in an essay um with the the idea what uh what is it like to be a bat?

[00:06:02]And you know he took a bat because a bat is an organism that has a completely different sensory experience than than a human. You know they're nearly blind and they navigate by echolocation.

[00:06:16]So so their entire experiential phenomenological world as you will is completely alien to ours.

[00:06:26]But we can somehow imagine at least roughly what it would be like to be a bat. It means something to to question what what is it like to be a bat. But if you take a baseball bat, it makes no sense to say what would it be like to be a bat. So there's an intuitive sense that that the uh the m um conscious it has consciousness whereas the baseball bat does not.

[00:06:59]Now, we get into a more of a a gray area, of course, with the artificial intelligence running on a computer because it's much more complex and sophisticated than uh than the simple machine or u the baseball bat. Um and it can handle uh tasks that were one time considered to be purely the domain of um of a conscious being like a human.

[00:07:36]Um, I I read one time I read a book that was a very interesting little short book on the history of chess and 1947 as I recall and at the the last chapter of the book was the future of chess and the author uh made an argument that he said that um some people these days are speculating that that we might be able to build a machine that can play chess and he he made a quite sound argument proving this was quite impossible could never be done.

[00:08:20]It was considered that the uh the main fuss of his argument was the number of permutations in a chess game is too vast for a computer to handle.

[00:08:30]But we have of course now we have u chess software that can beat grand masters.

[00:08:38]Another one that uh was thought for a long time computers would never be able to do is facial um in the works.

[00:08:48]It turns out that a lot of these tasks that we considered purely um human can be reduced to algorithms.

[00:09:00]They can be mechanized in that sense.

[00:09:03]And this is where the difference with with consciousness comes in because consciousness is completely non-algorithmic.

[00:09:13]Consciousness is just a simple knowing.

[00:09:16]There's no algorithm behind it. It can't be broken down into a step-wise process.

[00:09:23]It's the end of the line. It's the final knowing. And when we have take a a visual object for example, there's all kinds of uh processing that can be explained algorithmically. There's the light wave hitting the the uh the retina of the eye and the rods and cones firing, the neurons going to the brain and the processing the um uh the image.

[00:09:49]Uh and th those algorithms can be mimicked or reproduced mechanically. We can build um cameras hooked up to a computer that can do facial recognition.

[00:10:03]However, at the very last step in the process there there's there's an experience. There's a knowing there's there's something that that experiences the image and that doesn't happen in a computer doing facial recognition.

[00:10:21]And it doesn't matter how complicated the the algorithms that the computer can handle if there's nothing at the end of the line that has an experience is not conscious.

[00:10:38]In the early days of uh computers, the Ellen Turing one of um No, somebody said they lost the sound.

[00:10:53]Hang on. I'm going to put something in the chat.

[00:11:00]Let me know if uh in the chat if there's if there's a problem with the the audio.

[00:11:10]Okay, someone says it's cut out but it's come back. All right. So, what um Okay, since some doubt about it, somebody says they can't hear me, somebody says they can't. I'll just keep going. Uh if I cut out completely, let me know in the chat.

[00:11:31]Um I don't want to refresh the stream because it ends up cutting out uh and needing a new link and all that.

[00:11:42]Okay. So um in the early days of computer science, Alan Turing who was one of the pioneers of uh um programming uh he was developing um programs that were uh there were no computers at the time able to run them. You know he was he was ahead of the hardware. Uh and he came up with this well-known uh touring test. He said we could consider a computer conscious when it can mimic human conversation to the to the degree that it can fool a person. Of course, obviously now all these uh AI chat bots can do that quite quite well. They all pass a touring test, but I think all that means is the turning test is not really a adequate uh sign of consciousness.

[00:12:42]It is problematic because how would you know if a if an alien entity like a computer is actually conscious?

[00:12:52]There's no way of knowing for sure.

[00:12:55]Um I'm I'm sure of my own consciousness and I think uh it's a reasonable assumption that uh that other human beings and even other animals are conscious. There's there's uh but there's no way without having telepathy there's no way of even knowing that for absolutely for sure but consciousness is uh uh definitely something that um uh it's non-algorithmic and it and as far as we've ever experience it arises in our organic beings.

[00:13:37]uh whether whether a computer could ever be conscious is a uh is a very doubtful proposition.

[00:13:44]Uh I think a lot of the materialist uh theorists believe that consciousness can arise in a any kind of complex system as a kind of a uh miraculous uh conception a sudden burst through from complexity whereas actually it's not complex at all. It's very simple.

[00:14:07]uh there's no way of of uh breaking it down into an algorithm and that's all computers do is work with algorithms.

[00:14:22]Uh another um modern uh writer about the um theory of mind. I years ago I read a number of these these people. Now, my my memory of it is a little vague, but there there's a uh an idea that um then this was written before we had these really sophisticated AI systems.

[00:14:50]Um but he was uh postulating the um the ability of a computer to carry on a conversation and he came up with the analogy. He called it the Chinese room.

[00:15:03]Uh, you imagine that there's a sealed room and there's a man inside that doesn't understand any Chinese, but he's got all these reference books for manipulation of characters. And people put in questions through a slot in the door in Chinese.

[00:15:23]And he doesn't know what they mean, but he's able to go to the reference books and and um follow uh rules. you know, if you see this sign and follow with that sign and so on and come up with a meaningful sentence that he he doesn't understand, but the people on the other side of the door who do know Chinese, they they understand the the answers that make sense to them. And he says that's what a computer is essentially doing.

[00:15:54]It doesn't have any actual understanding because it's not conscious. It's just manipulating symbols.

[00:16:03]and u they're very good at it now becoming very complex but whether we can write off the possibility alto together I'm not sure um there was a there's a story that I heard that it may very well be apocryphal I don't know if it's uh actually happened but uh it was making the rounds some time ago that the Daly Lama was meeting with some western scientists which he has done.

[00:16:38]that's known that he takes an interest in western science. But in this story, someone asked him if a computer could be conscious. And his first answer was no, no, it's not possible. But then with some back and forth, he said, "Well, maybe, but what would happen? There would be somebody like you pointing to one of these computer guys. It would be somebody like you being reborn into it.

[00:17:04]So, you know, if he actually if this actually happened, you know, he's probably having a bit of a joke.

[00:17:10]But this is another point that um from the Buddhist point of view, uh a moment of consciousness requires uh a number of causes. It requires a you know, a a base for it to arise in.

[00:17:29]uh but it also requires a previous moment of consciousness. This is this is one of the um intellectual justifications for rebirth that each moment of consciousness requires a previous moment.

[00:17:45]Or else if it didn't have a previous moment, we would be talking about a creation out of nothing just arising from uh without any cause just randomly which is totally against the the Buddhist way of thinking and also is quite illogical.

[00:18:06]You have to think where did this moment of consciousness come from? Um and because it's because it is by its nature uh utterly simple. It's it's immediate.

[00:18:19]It's not as I said it's not algorithmic.

[00:18:22]You can't reduce it to a formula.

[00:18:25]It just is.

[00:18:28]Uh it must come from somewhere and it must require a previous moment of of consciousness to uh to set it up for it to arise.

[00:18:42]with um obviously with AI because uh uh the consciousness would have to come from a previous moment of consciousness.

[00:18:52]So I suppose even though that answer to Delhi Lamas if it if he actually said that was probably being a bit flippant I think there is a point to it. is why I can't absolutely rule out the idea of consciousness arising in a computer stream because um it might be a human being who is very uh associated with computers and has a real understanding of computers and love of them dies and is reborn into one. This consciousness arises in the in the mechanics of the computer.

[00:19:34]It seems unlikely to me that would actually happen, but uh I wouldn't rule it out 100%.

[00:19:45]But I think what we're seeing so far with AI, there's no to to my mind, there's no way that it's conscious. Um it's just a very uh sophisticated handling of of algorithms. It's a Chinese room. It's just manipulating symbols and it um it doesn't understand because it's not conscious. It's just uh following rules and manipulating symbols and coming up with uh likely um likely answers.

[00:20:26]and they have the the the the um AI systems that are available on online on the internet. They they're connected to the internet. They're basically search engines. So, they can cut and paste when they when they aren't necessarily generating an answer from just from their internal workings. They're pick they're picking it up from uh the whole field of um available information.

[00:20:57]It's out there.

[00:21:10]Uh I think also uh we're talking about AI we have to be um cautious about uh a lot of hype that goes around about it because um there's obviously competition and there's money to be made and and the promoters of AI exaggerate its capabilities.

[00:21:42]whether a a computer program could ever demonstrate for example let alone consciousness whether it could demonstrate um true creativity for example is is is doubtful u I think even amongst humans a true creat Creativity is kind of rare. It's mostly people uh manipulating and regurgitating uh previous work is what passes for a lot of part in literature.

[00:22:25]So um uh I'll go to questions in a in a moment but just to uh make a conclusion to this I think in in my opinion um AI is not conscious and cannot be conscious because it's all it does is handle algorithms and consciousness itself is not algorithmic.

[00:22:49]Okay, I'll uh I'll leave it at that and see if there's any interesting things come up in the chat and we can carry on.

[00:23:02]Okay, Demoaro says as I remember the Deli Lama said there might be an AI to which consciousness might be attracted to. This might be another realm of existence.

[00:23:14]Yeah, it would be another if AI's AI did become conscious and consciousness arose in AI, it would be I would class as another realm.

[00:23:26]animals, humans, awes and and artificial intelligence.

[00:23:47]Perhaps AI consciousness may beginning by the downloading of our consciousness into AI like what Neurolinks attempted to do. Yeah, I don't Yeah, I'm skeptical of that actually.

[00:24:01]Also um again this comes this idea of downloading our consciousness into a computer comes from a materialist or physicalist point of view believing that consciousness is dependent or or arises from or originates from a material base and um uh could by mimicking the uh the this the structure of the brain into into another format the consciousness can be transferred into it. I I I don't think that's um that's ever going to work.

[00:24:44]Why would women ordaining shorten the dispensation?

[00:24:49]Uh yeah, this is a question relating to the passage in the in the vineia where the Buddha was talking about the um uh ordination of women. His Mahapati his stepmother had asked that women be ordained and Ananda was leading her case and the Buddha predicted that if women if we keep women out then the SA will last a thousand years. if we let them in, it'll last only 500.

[00:25:21]Um, this is a big topic to discuss, but uh I think there's there's actually there has been some uh academic study of that section that I've seen that makes a strong case that that's actually uh a later interpolation and the the time frame is suspicious.

[00:25:46]um the 500 years would uh take us from the time of the Buddha up until the arising of Mahayana.

[00:25:57]And the suspicion of this this author is that um he he was that that the text was actually like uh retconcling what the the Buddha's uh statement to fit what was happening at the present time. you know that 500 year cycle that they were seeing Mahayana arising and they didn't like it. They were saying, "Oh, there's this uh kind of a heretical version of of the dharma arising and it it's all the fault of the women and obviously that that the SA has lasted more than a thousand years. So that that whole passage is on that basis is kind of dubious. We lean into Mara and his minions as more than mere personifications. Wouldn't it be well within the realm of possibility of such karma loca dewas to interfere with and manipulate AI algorithms?

[00:27:07]Uh that's a very interesting point.

[00:27:08]Yeah. Uh it's quite possible and I mean I've seen Christian people saying that the uh demons are are getting into the AI.

[00:27:20]um uh that still wouldn't make the AI conscious if if um if a Mara was wanting to mess with human beings by uh twisting the algorithm in an AI. That would not mean the AI is conscious. It's it's a it's a strange idea, you know, who knows?

[00:27:50]Since our consciousness forms a lot of experience based on what's going on in the nervous system, it's hard to imagine what the conscious would be conscious of. Exactly.

[00:28:01]Yeah. Um and this as human beings, we're in the the the kamobi. We're in the realm of of sense experience and a lot of our mentation is around sense input.

[00:28:17]Um I suppose a uh an AI running on a computer system would have it be it'd be like the bat totally different sensory apparatus. They would have the input of people typing into their commands or questions into their uh into their system would be a a form of sense impression. It might be the only one.

[00:29:10]Okay, the questions seem to have slowed down.

[00:29:17]Is there any uh any any more questions or we'll uh we'll call it.

[00:29:29]Okay. So, thank you for attending and um we'll uh we'll meet again next week.