Overhyped or Deadly - Should We Be Afraid of AI and What’s Coming?

[00:00:00] Speaker A: What is personhood? What is sentience? What is intelligence? Do we have something more to us than just the neurons firing in our brain? If I can replicate that in a model, is that consciousness? These are hard questions. [00:00:09] Speaker B: So the privation of being, the absence of being, in the Augustinian line of thought, that means the absence of good. [00:00:16] Speaker A: Good, yeah. Is evil. Yep. [00:00:17] Speaker B: It raises an interesting question for him about whether even just our dialoguing with it is, in a sense, is not a healthy thing. [00:00:24] Speaker A: Do you honestly think that's Aslan? Do you really think that's a line? Or do you think it's the donkey? It's a hype cycle. And the hype starts to really build. Oh, my gosh, this is going to be everything. So you end up in a trough where all the headlines will be saying, this is a waste of time. It's going to deliver nothing. We're now starting to get, you know, singularity type capability coming online. What do we want the future to look like? What do we want to shape it towards? Because if we just leave things on autopilot, we might not like where we land. These questions that people are starting to wrestle with, where do I turn for answers? And we're sitting here with the answers. I love that in your podcast that you start them with, you know, everything that's good, it's his. Everything's rubbish. That's all me. That's all on me. And may you, like, forget it. By the time you log off, I'm like, that's great, mate. It's really good. [00:01:11] Speaker B: Well, I said to someone today, I'm a mad Irishman, so there's plenty of rubbish comes out of this. Actually, speaking of rubbish, let's start there. Did you ever have a phase? And I'm asking you this because I saw the Daily Wire guys asking each other this yesterday. Did you ever have a phase where you thought there was a conspiracy theory might be real? Was there ever a conspiracy theory that captured you? [00:01:34] Speaker A: No, is the short answer. I remember having a conversation, actually with some colleagues about the moon landing. And the conversation was, I was using it as an example. What I was trying to point out is that when you have people inside of a large organization, they're going to have different viewpoints. So I was like, you know, like the moon landing, for example. I mean, there'll be some people in our organization just based on numbers, who don't believe that we land on the moon. And one person at the table goes, yeah, but we didn't land on the moon. And then what happened was this big discussion for like, you know, 15 minutes of everyone at the table trying to convince the person that we land on the moon. And when it all settled down, I was like, thanks for proving my point. Like, what difference is that? But for me, I don't know. No, there's never been one that's sort of risen to the top. JFK never asked some questions about it. [00:02:15] Speaker B: I don't know. [00:02:15] Speaker A: Titanic did it really, like, you know, UFO Titanic? Yeah, yeah, yeah. Area 51, maybe Area 51. When I was little. When I was more into like sci fi. My teenage years, maybe. I don't know. [00:02:26] Speaker B: Well, that's what got me thinking about it was. Cause the disclosure. [00:02:28] Speaker A: Yeah. Of the new UFO coverage. [00:02:30] Speaker B: Yeah. Which is just hilarious. Cause all you're gonna end up with is just a slightly more documented version of what you already have. That's I guarantee you. But what's interesting and it seemed relevant for our convo tonight is straight away, the AI slot arrived. In the last week, there has been so much fraudulent images, fraudulent video. Now, some of it's really obvious, only probably the, you know, the uninitiated will [00:02:54] Speaker A: be fooled by it. [00:02:55] Speaker B: But some of it is compelling enough to make you believe that what you're seeing is, oh, this is a real photo. [00:03:00] Speaker A: Yeah, yeah, yeah. [00:03:01] Speaker B: That's not. It's not. [00:03:02] Speaker A: I think there's something about. If you look at all the good quote, unquote conspiracy theories, the ones that seem to stick, there's a kernel of something in there, Right. So even Area 51, you go, okay, well, there is a secret. You know, like the military has a history of withholding information until it's no longer strategically relevant. So there's like a kernel of truth or like a possible. It has to be. If the entire thing was improbable or impossible or preposterous, then people wouldn't pay attention to it. So I think when you look at some of the AI being able to amplify, being able to augment, I like to say it's augmenting our intelligence. That also means that it's going to augment or amplify not necessarily the best parts of who we are as well. [00:03:43] Speaker B: And what really frightens me about AI is actually its potential to analyze human vulnerabilities in that regard. And go, this is the most effective way to deceive a group of people. This is what you say. And, you know, once it really gets applied in that space, it will be a very powerful propaganda tool because you have images that seem real and a crafting of Things that are just. We're vulnerable to. [00:04:07] Speaker A: Yeah, I think. I think the amplification thing really plays there. Right. Because if you think about it, if I really want to. If I wanted to work on some propaganda, if I wanted to try and talk with psychologists and talk with anthropologists and try and figure out what are going to be the triggers that are going to get people to engage, I could do all of that today. I think the ease of access that is provided through augmented tooling, so AI tooling, I think it's still pointing towards humans giving direction, but in the same way that if you kind of imagine before the Internet, I could go get all the information, I just have to go scour the world's libraries and jump on planes and whatnot. And now I could just do it with a web search. It's just reducing that last mile of access. Does that then open up democratisation of capabilities to nefarious actors? Yeah, well, it does. Alongside with the good as well. Yeah. [00:04:52] Speaker B: I want to talk about. I want to come back to that point about democratization of the technology. But before I do that, rather than probably what typically happens in these conversations, it's either it's all doom and gloom or this is the most amazing thing and you should just be mainlining AI right now, otherwise you're an idiot. And often there's no balance. So let's start with what are some [00:05:13] Speaker A: of the positives here? [00:05:14] Speaker B: It's not all bad, that's for sure. [00:05:16] Speaker A: I think even taking that approach, I think is a really good one, which is, say, look, what are the bounds when we're thinking about the future? And a lot of this is what I sort of do as a day job, right. Is looking at the future through the lens of, is it something which you can predict, anticipate, or is it something that you're actually trying to shape? And I think that some people, when you think about sci fi, you know, we're both, you know, movie fans. Name your favorite utopic sci fi movie. There aren't any, like, Star Trek's the closest. Star Trek's the closest. And what does it still tries to bring in war and the fight with the Romulans and everything else to say. There's. There's all the Klingons originally, because there's tension, because drama is what's interesting and Utopia is kind of boring. [00:05:54] Speaker B: Yeah. [00:05:55] Speaker A: Name your favorite dystopic sci fi. We're going to have a long. We'd spend the whole conversation talking about it. Right. And I think when we think about the future of any kind of technology we skew either utopic or dystopic. And those are the extreme boundaries. It's going to be the end of everything. It's going to be the beginning of everything. And if you look at any of the advanced information technology of at least our lifetimes, and before it's somewhere between the two. Right. It comes back to the purpose. So I think that in fact, the very definition of the word technology. [00:06:23] Speaker B: There is a distinction here though. [00:06:24] Speaker A: Right, go on. [00:06:25] Speaker B: In the sense that AI is, it's not like typically sort of benign technologies we've known in the past where you really do require a human actor to actually be a lot more actively involved. Like a motor car, you have to have a driver taking it somewhere before it becomes a tool for bank robbery, an axe, before it becomes a murder weapon. It has to be wielded by a human agent. Because AI has the ability just to sit there doing things without. Once it's set in motion, the human intervention is required as the first domino kick, right? [00:06:56] Speaker A: Yeah, that's right. [00:06:57] Speaker B: But then it can actually engage with the ecosystem, if you like. [00:07:01] Speaker A: Yeah. I think that there is a different kind of access that's enabled by the nature of the interconnected Internet and being able to say to an agent or an agent capability, you now have tool use. You can now go use these things. You can do these things. Now go do them. I think one of the things around planning, which is a tricky thing in the AI world, being able to plan out and say, hey, look, you've got an objective, go achieve it within side these bounds. But how you achieve it, that's quite different to traditional programming where you would say, I want to achieve this. I'm now going to give all the instructions. So that I can see as a difference. But it is still, I mean, we can get into Aquinas. This is still a first mover. There's still a human who's saying, actually there's a goal that you have here to try and achieve. I think the thing I was going to say about technology, the definition of the word technology, going back to the Greek, is like the practical application of scientific knowledge. So when you're applying scientific knowledge to a practical purpose, the question that immediately follows is, to which purpose? Yeah, what is the purpose? And so if you're pointing it at a purpose that is true and good and amplified, then you can amplify the best parts of it. So some of the great use cases of the current flavor of AI, particularly large language models, which is what's got everyone's attention over the last couple of years. Ignoring the previous 75. Focusing on the last couple is pattern recognition, which is a huge thing, I think understanding of human language. I've said this to a few people, understanding human language is a heck of a win. I'm old enough to remember using Dragon Naturally speaking 1.0 and you had to train it. You had to spend 45 minutes reading [00:08:30] Speaker B: these and even then you talk slow [00:08:32] Speaker A: or you had to put on a terrible approximation of an American because they couldn't understand Kiwi. From there to. I can have a chat with one of the. Even the. Even the tool that everyone talks about that kicked it all off with ChatGPT. I can talk to it in English slang, kind of spelling mistakes. I can talk multimodal. I can chat with it in voice and it can understand what I'm saying. Not well, we can get back to that. As to what it does understand, it can take my words and it's not having to be in a formula. I don't have to say do this if this then that. That's huge. That's a game changer for interfacing with the machine. [00:09:05] Speaker B: And that's the thing is there's great power. There's. But there's also an illusion that there is a shell game happening and we're very vulnerable as human persons because it's mimicking something like you say it's not. It's. There's an anthro. More anthro morphizing. We're giving it an. Anthropology doesn't have anthropomorphizing. [00:09:27] Speaker A: That's what I was searching for. [00:09:28] Speaker B: And there was. There's also a personification but it's not because you're talking to it, you know, it's speaking back. What you've got is a very clever algorithm that is doing a lot of math very quickly and very, very efficiently. [00:09:43] Speaker A: Yeah. So there's a few things to unpack there. Right. If you look at the. You can go back to the attention is all you need paper that sort of kicked off the transformer. So for those who aren't big AI fans or AI know the GPT part of chatgpt stands for generative Pre Trained Transformer. There's a whole bunch of changes in the nature of how the architecture of deep learning works to be able to understand human language, train it on the entire Internet, including Reddit. What could possibly go wrong? Right. But it is predominantly. And I mean there are people who are far more like who are true experts in this field, who can explain it better than I could. But, you know, I've had the great opportunity and blessing to be able to work with a number of these kind of real leaders in the field. And when they dumb it down for me, you know, one of the examples I'll use is if I said to you the cat sat on the, you would say hat, Right. So a lot of people would go mat because of the Dr. Seuss, the cat sit on the mat. But then the fact that you go hat is a perfectly. It's in the valid choices. The chair, the rug, the table the cat can sit on. A whole lot of things. It depends on context. What the models are doing is based on their training data. What's the most likely, the most probable solution? The cat probably sat on the mat if it's based on the training data. Right. What we get learned is context. And we learn it from our own learned experience and growing up and everything else. The models are just exceptionally good at probabilistic determination, and that's something that we're not good at. The evidence that humans aren't good with probabilities is the fact that people still buy lotto tickets. Right. The odds are not in your favor. The odds are never in your favor, but we still do it. And so I think that we're used to asking our computers, what's one plus one? And they calculate that the answer is two. We ask these models, what's one plus one? And they say, well, the answer is probably two, but it could be window, it could be 11. It kind of depends on context. And what happens to us, I think, is that we are fooled by the fact that in the past, through basically all of human existence, the ability to speak, to bring together thoughts, to have conversation, is a proxy for intelligence. So then when you see something that is able to speak and construct seemingly fluent phrases, we immediately default to. Because as we have for generations, go, well, that must be intelligence. Now, we know that it's far more complex than that, But I do think we get, as you said, it's not. Well, it is maybe as fooled, but it's also that we're drawn to that. [00:12:02] Speaker B: Yeah. There's something we know instinctually that we're seeing. There's a mimicking of that thing. [00:12:06] Speaker A: Absolutely. [00:12:06] Speaker B: And also I have this thesis about AI that if it had arrived in its current manifestation prior to the development of social media, so let's say it arrived about 20 plus years ago, I think it'd be a very different scenario. I think social media has made us A lot more vulnerable. I think it would have. We would have just looked at it and thought, wow, this is an amazingly efficient machine. It's phenomenal. We might use words like clever and all the rest of it, but I think there's a distinction now because of social media and that as a vehicle, a channel for it to then sort of merge with and create all sorts. [00:12:44] Speaker A: So let's build on that. If you think about what social media has done, and I know you've spoken about this quite a bit in the past as well, it's removed or reduced human interaction to quite a low fidelity experience. Right. You show the best and hide the rest. Originally it was 160 character limits. So you were constrained the full human interaction and experience. All the messiness and the organic and the emotional is reduced to this far thinner, low bandwidth interaction. And as part of that. And we were similar vintage. So we remember. I remember thinking when Facebook first came out, I'm like, oh, this is gonna be interesting. Cause people's real names are connected so they're not gonna be attacking each other. Which prior to that in kind of gamer culture, it was a pseudonym, chat rooms were. But exactly the opposite happened. People didn't seem to care that their name was associated with it. And people would again reduce and they would attack and they would go to the hashtag and all this kind of thing. So I think we become more susceptible to something that can engage on those same terms when we've reduced the barrier of entry, if you like, to the greatest level. The flip side for me on that is my take is I think that it's interesting that particularly this latest widespreading of the capabilities happened after Covid. And I say interesting and I mean more from interesting from a divine timing perspective. Because it's almost like with COVID we generally had a shared experience of what it is like to be removed from. You had your bubble, but for the most part you were removed from society. If you're engaging with anyone else, it was through artificial means. Yeah, zoom calls and everything else. [00:14:25] Speaker B: Yeah, that's right. [00:14:25] Speaker A: Right. And so initially we were like, okay, this is cool, I get to go to work in my pajamas. And then after a while, even the introverts among us were like, I don't know if I like this. Having the entire experience reduced to just a screen. And then particularly if you were, you know, if you're Catholic, for example, it was quite constraining because we couldn't actually physically get together and physically gather. So we come out the other side, in and out in and out, in and out. We come out the other side mostly of that. And it's almost like. I'm not saying that God designed it, but it's like we have this experience. This is what it would be like if the only way you interacted with other things was in an artificial means. And then we unleash on ourselves this new capability, which sounds like. Looks like, but it's entirely artificial. And I think it's fascinating to see that we always have lost the lessons from the COVID experience of what it means to truly cherish the full human experience. [00:15:14] Speaker B: The embodied reality. [00:15:15] Speaker A: The embodied reality. And so you're left with this shadow. Someone asked me about, you know, personhood and all this kind of thing, and the Catholic in me wants to say, well, I can never engineer a soul. I can't code a soul. And if my inherent dignity being made in the image of God comes from that, I don't have any worries about this thing hitting sentience or even being a thing. I do think, however, people can be drawn in, attached, fooled. And so that's something which, as a society, we've got to have these kind of conversations to go. Hang on. Let's just reset some of the rules. Speaking and constructing a sentence is no longer a valid proxy for intelligence. So we need to establish new ones. [00:15:56] Speaker B: Well, the Turing Test doesn't work. No, it's very apparent that the Turing Test was very much a thing of its time. [00:16:01] Speaker A: It was also positioned in a different way. I think when you look at. I mean, Alan Turing's paper in 1950 is kind of where you sort of see the concepts of machine intelligence. And you've talked about, you know, the Dartmouth. Dartmouth Conference, where it all started to kick off at that time, it was the sense of, like, well, if you can get a machine to do something that is indistinguishable from a human or seems to be acting in the way that a human would act, that feels like a pretty good measure. [00:16:25] Speaker B: And they were calling that intelligence. That was probably the first mistake. Really? [00:16:28] Speaker A: Well, yeah, I think it was potentially just a simplification. [00:16:31] Speaker B: Well, it's also probably a product of a mechanized view of the world that it also had a mechanized view of the human person. [00:16:38] Speaker A: Yeah. To be able to reduce to a point of saying, well, how do we define intelligence? There are multiple dimensions to it. Let's define it on this dimension, and if we can kind of get that. So I think today it's rewrite the rules. And I do feel that society having more of that conversation professionally, personally, spiritually, we need to be having these conversations in a different way than we have before because now we have something that goes those traditional rules, the measures, the cognitive sort of shortcuts, they don't apply. You've got to reset. [00:17:06] Speaker B: We didn't really get to what the positives were saying. What do you think are some, I guess, benefits and also maybe where do you think its biggest potential for human flourishing? [00:17:16] Speaker A: So I think so in understanding language. It also means understands multiple languages. So the ability to translate. How many things get lost in translation that now no longer become lost in translation. What's the opportunity that opens up through language to different interfaces, to people who are differently abled, but also in terms of actually changing us away from the screens and the lock in and the zombie. If I'm using that because it's the only way I can access useful information alongside useless information. Can my kids tell the difference between me reading the paper, paying some bills for the family and doing a crossword puzzle? I used to be able to tell the difference when my parents were doing those activities. For now it just looks like the same block of glass. [00:17:54] Speaker B: Yeah, it's the same thing. [00:17:54] Speaker A: What if that moves away? So the more ambient compute more of the actually can you pay those bills and have them sorted? And I'm not looking at a screen when I do it. So I think there's an opportunity to break some of what we've as a species kind of formed into Homo mobilis [00:18:07] Speaker B: or even to the potential for tools that will, like we've got various tools now that will try and limit your screen time or brick your phone, but a tool that will actually adjust to your own particular vulnerabilities and. [00:18:21] Speaker A: Yeah, and I think adapting also to the use of why are you using a computer or using a phone or whatever else? How does that start to change if you can have a certain amount of automation helping that in the background. So I think when we say these technologies, I think the large language model for me is interesting because it's understanding of language, I think very useful with research. Now, if you think, if you think in terms of there's a whole conversation to be had about novel research versus, you know, can you really have any novel research come from these models? As a, as a species, we're producing so much discovery and research. And is it possible, even with the World Wide Web, which was initially created to connect universities, to connect and share information, is it possible for a researcher to truly read all the research? Right. So if you think about, you know, the examples that have come through around some variations of these models being able to help with diagnoses. And it's like because it can read or can consume and I don't even like saying it because that's an anthropomorphization, but you can create a model that has read all consumed all of the oncology research, all of the pharmacology research [00:19:29] Speaker B: and the amount of scan data, that's why it's speeding up the diagnosis for breast cancer. It's got so much human data to deal with. [00:19:35] Speaker A: You know, really a good friend of mine, his background before he actually became an expert in AI was pharmacology. And so the ability to, you know, the things that have been unlocked around and some of the stuff is not associated with large language models. I'd point out when we AI, it's become synonymous for most people with the recency bias around generative AI and specifically large language models. But you know, protein folding, these kind of things that can help with creating new drugs that can be well off label use. [00:20:03] Speaker B: For me, that's where I think I see some of the most amazing potential. Yeah. Because the companies are not going to pay for it and it's time consuming, it's costing, there's not much benefit for them off label use. But having an AI that can scan all of that and say hey, here's some potential off label that becomes life saving. [00:20:19] Speaker A: Yeah. So being out, being able to run simulations, being out of. There's all these kind of areas where on the edge of what is possible with these capabilities which you start to say, well gosh, if we could solve that and we could solve that, then that could really shift the problem from where it is at the moment to where it could be. I think there's opportunities around embodied AI or physical AI. So robotics has been held back a lot by the fact that you have to pre program the entire environment in which you are sending a humanoid robot into or any other kind of robot, being able to have models where the deep learning can then enable them through senses to anticipate and adapt and respond to the environment is a whole separate conversation about embodiment being again a proxy for intelligence or humanity. But I think that that's really interesting and exploration and there are, I think there's myriad others as well. You know, just being able to put ourselves in a position where the nature of how we use the tech. So even think about vibe coding today, the nature of how I can speak into being an element of software. Now as a software engineer by background, I would be cautious about where that's used, but it can definitely augment and extend people's capabilities. Yeah. So what used to take me this amount of time now takes me this amount of time. Great. I can then move on to all the other stuff that I wasn't doing. And, you know, there's a thing called the lump of labor fallacy that we're just going to automate all the way the work and then we'll be done. And you go, that's never happened. We always find more work. There's always more work. [00:21:42] Speaker B: Well, this is an interesting question for me. Well, first of all, I think. Love to hear your thoughts on this. I think AI has absolutely been oversold in the last 12 months. Its capabilities, I think. So under delivered, over, promised, probably because you had people who were getting all sorts of startups going and they were desperate for investors and customers. That's my. [00:22:01] Speaker A: Yeah. Are you familiar with Gartner's Hype cycle? [00:22:03] Speaker B: No. [00:22:03] Speaker A: Okay. So in the IT world, Gartner, I mean, people are familiar with Gartner, but one of their famous models is called the Hype Cycle. And so they've used it for a number of technologies. They've used it, you know, decades. And basically the Hype cycle have a slide to show it, but in effect, it looks like it starts with the innovation trigger and then tracks up to what they call the peak of inflated expectations. And then on the other side of the peak of inflated expectations is the trough of disillusionment, which is lower than the original innovation trigger. And then on the other side of that is the. Not to oversell it, but it's called the slope of enlightenment, which leads into the plateau of productivity. So the curve kind of goes like this and comes down and comes up like that. And the point is, it's not a maturity curve because the technology can mature at whatever rate. It's a hype cycle. So often what happens is a new technology comes along, an innovation trigger happens, and the hype starts to really build. Oh, my gosh, this is going to be everything. And if you want a case in point, if we're having this conversation four years ago, we're talking about cryptocurrencies and NFTs. Oh, yeah, exactly right. The hype gets high. This is going to change everything. And then what happens is the realization of where it is right there is not quite at the same level as hype. So you end up in a trough where all the headlines will be saying, this is a waste of time, it's going to deliver nothing. And then over Time, if the technology survives that hype crash, what starts to see happen is we see use cases emerge and then you're enlightened into the plateau of productivity. So an example I would use is if we're taking cryptocurrency and all of that distributed ledger technology, which is what underpins things like the blockchain, you go, okay, it actually has applications, some genuine productivity, and things like decentralized digital identity. Now, that is thoroughly unsexy as a topic. Right. So it's not very hyped. But if you think about what was promised and then what was like, it's never going to be able to do anything. There is actually value that comes through. And so when you look at this family of AI technologies, what you're left with, I think, is what's referred to in the industry as the capability overhang. Just the capabilities that have been delivered so far. We're not even fully utilizing to the max. So there's a lot of capability that is on offer. But there are plenty of people who are saying, yeah, but wait, just wait, wait until we get to AGI General, artificial General intelligence or super intelligence. It's going to be incredible. It's going to be amazing. And I think some people are naturally drawn to ramp up the hype. Sometimes they might have incentives to do that. Others are naturally drawn to drag it down. Whatever it is, it's going to be a waste. It's going to be terrible. There's those polar opposites like you talked about before. But inevitably with any technology, it's a tool. And so the ability to turn it into a tool, preferably rather than a weapon, is to be able to bring it into a stage where you go, no, I can see some productivity from that. [00:24:39] Speaker B: Yeah. And I think that the thing for me too is that probably that hype curve also applies to the fear curve, I'd say as well. [00:24:46] Speaker A: Yeah. [00:24:47] Speaker B: And there are, we'll talk about this, but there are definitely issues with AI that we should be wary of. But it's also clear, I look at it, I think, well, I was just talking to someone recently, couple that we know, really good friends with my wife, and he works for one of the Federal Reserves, one of the states in America. And he said, in actual fact, the conversation is not really about AI, it's about quantum computing, because that poses a much more serious and imminent threat to security online. Because you can crack anything, everything becomes crackable and very almost instantaneously. So how do you actually secure banking when you can crack something that quickly? So that's more of a warrior for them. I look at AI and I think, yes, AI in and of itself, sure. But really it's what you couple it with that starts to become the issue. So one of the interesting things that's actually happened in the last. Very recently, actually in the last couple of weeks is that Japan now, for the first time since World War II, has started manufacturing and selling arms again. And that was never a thing previously because of World War II, obviously, they embarrassed themselves, shall we say, and were told not to do that. So now they're selling and their that robotics, that's the bread and butter. And it's drone technology. Now. Japanese drone and manufacturing capabilities combined with AI, you got something that potentially is a whole frightening thing, AI on its own, not really so much. [00:26:08] Speaker A: Most technologies, I think, are more powerful when seen in their combinatorial state. So, like I, you know, I remember giving talks over the years about how technology isn't by itself an innovation, it's a catalyst for innovation. So the fact that I can combine a different business model and a different profit model and a different marketing idea, and I can speed those up through the use of social media, or through the use of mobile tech, or through the use of AI, or through quantum. It's their ability to combine. And particularly when you combine catalysts. I wasn't very good at chemistry, but if you have catalysts, they can speed up the whole reaction by taking out some of the energy required to get the result. So that's kind of a rushed version of it. But if you think about these technologies as being something that we can connect to. Okay, you talk about in terms of Japan. So because of the demographic collapse that they're facing into. The reason why one of the key drivers, in my opinion, for the massive success they're having around in robotics is because they need to be companionship bots, but also aged care. Yeah, they don't have enough people to actually help with nursing. Now, where do you use. There's a question, right? Where do you use that kind of capability? Do you use it in helping, you know, flip beds and open doors and things like that, or do you use it in place of human empathy? Obviously, I skew towards the former rather than the latter, but it's the same sort of thing. At what point do you say, well, bringing in artificial intelligence capabilities, bringing that in is going to help me create a more rounded product or solution. Okay, what is that product or solution? Because if it's a weapon, that poses a whole bunch of moral challenge. If it's Companion that's a whole bunch of moral challenge. And so all of this comes back to there are no solutions, there's only trade offs, which is the Thomas Howell quote. [00:27:47] Speaker B: Yeah, yeah, it really is interesting particularly I think about warfare. There are big implications beyond just the frightening potential to have completely artificial and very efficient technology, lethal technology at work. But there's also the factor of that has a bearing on now things like just war doctrine, which has survived for thousands of years and even in modern warfare, while there have been questions like nuclear weapons do change the game. I think missile technology as well. But with AI involved now there's a whole nother component that really does completely require a re examination of just war. [00:28:28] Speaker A: I mean I think, in fact I remember many, many years ago with a very young and bright eyed and bushy tailed. It looked very similar to this unfortunately Brendan, when we were talking about just war doctrine. And I remember might have been Iraq war, wasn't it? No, might have been. I remember because of John Paul II, Pope St. John Paul II. I'm going to misquote him here, so you have to look up the actual quote. But he was talking about is it possible in a world with atomic weaponry to have a just war? So his point was could an atomic war or a nuclear war ever be a just war? [00:28:58] Speaker B: And the answer is no. [00:28:59] Speaker A: And the answer is no because it's just the sheer amount of it. So his point was saying even at that point, this doctrine that you're right has held true for centuries. Suddenly in the space of decades or years was being fundamentally changed. Biological weapons, same question. So I think that we look at where we use, I said that tools or weapons thing for a reason. Father Paolo Benanti, who's the AI advisor to the Pope. So he talked about that, you know, these things can be tools and weapons and it's a choice to us whether we want the plowshares or whether we want the swords. And I think that's the challenge is to say just because it can doesn't mean it should. [00:29:35] Speaker B: That's the, I guess that's the, that's a good place to talk about. What do you think of the, the dangers and the threats? Because I see there are certain, I look at AI and I see there's potential harms that are not yet realized that very easily could be. And there are, I think, shall we say, intrinsic arms that come with the territory of what AI is. The nature of how we are interacting with Inheritor interacts with us in a sense, you know, there's a And it's not really so much AI's fault, in a sense. It's a. Again, it's a chain on a long disassociation of our own personhood from the way we interact with the world around us. But you can see clearly there are people who want to wield this power and have all sorts of utopian fantasies, really, about what they will be dystopian, ultimately, and they think it'll be in our best interest. But, yeah, I'd love to hear your take on what you think. What are the harms? How do you read that? [00:30:28] Speaker A: Yeah, in fact, I've written a little bit about how, because I got asked once on a panel, what's my biggest fear with AI? And I replied, sleepwalking. So my biggest fear is sleepwalking. And what I meant was, if you look at what sleepwalking is, it's sort of not being aware and stumbling forward. And so I think if you think about AI, it's actually with any technology. But I do think AI is the topic du jour. My big fear is that people don't stop and ask the big questions because the questions themselves are hard, let alone answering it. So we talk about, you know, what is personhood? What is sentience? What is intelligence? Do we have something more to us than just the neurons firing in our brain? If I can replicate that in a model, is that consciousness? These are hard questions. And so my big fear is people kind of go, it's too hard to even ask the question. [00:31:16] Speaker B: And there's no institutional authority. Previously, you would have said, okay, the church is there, and they have 2,000 years of wisdom and deep philosophy and theology to engage with it. Well, that's not institutions, although what you [00:31:26] Speaker A: are seeing with things like the Minerva dialogues and others, you are seeing this drive towards actually a rediscovery of. Hang on a second. Maybe the great faiths of the world have this different area of expertise, and maybe people, you know, like Father Benanti is a good example, like the professors in the area, and then they're bringing in theology and morality and these kind of considerations. So I think my big fear is that we don't have the staying power. I listen to one of your podcasts recently where you're talking about, like, if we stop reading long form, if we stop being bored, if we stop contemplating, then it atrophies. Right? The ability to do that drops. I worry that to wrestle with some of these questions about should we not could we anymore, but should we. To wrestle with that requires a different level of Investment of time and focus and if people are asleep at the wheel, then that's a problem. So this is why I think the conversation is important because it's waking people up in every sphere of life to going, hey, as a species we're now starting to get, you know, singularity type capability coming online. What do we want the future to look like? What do we want to shape it towards? Because if we just leave things on autopilot, we might not like where we land. So that's probably my biggest fear. [00:32:38] Speaker B: It's interesting because I think even the comment you made earlier on about work and we've always found more work, I'm a little bit more, I don't want to say pessimistic, cynical or suspicious of that claim. And I think you're right. In the past I look at the printing press or the motor car or, you know, the chainsaw will take away the job of the axe wielders. Yeah, well, no, in actual fact those technologies improved and there was always going to be a whole lot of other associated gainful employment around that technology, sharpening of tools, engineering and then also maintaining avengers, all that kind of stuff. Right. This time I think there's a bit of a difference potentially because we're now in a space where we're talking about the machines, looking after the machines effectively. It's got to the point, I think with someone like Elon Musk, whatever people think of him, where his whole universal basic income idea is now being treated in some quarters at very high levels too, quite seriously, as if we'll all just be happy if everything's mechanised and automated and then we'll just get money, which I'm sceptical. Why would you. No corporations paying us that money and where are the government's getting it from? So it's a fantasy in that regard. But the human person needs the dignity of work and I think there's only so much work. I think there's an apex, there's a threshold you cross where in actual fact you can't just keep creating new industries around the technology taking on too much of, you know, whereas previously you can motor cars. Yeah, sure, they create other industrious work around them, but I think we're getting into a point now where the simple truth is they just. We're not going to have and need that many people to do looking after the machines and then the people. How many people will have the aptitude for that? And they shouldn't have to either. But what happens when you've got. I Mean, they're even talking about, the joke was, what, a year ago that don't worry, go into plumbing, because, you know. But now they're actually talking about the possibility of robotics doing that kind of work. [00:34:30] Speaker A: Yeah. So I think that with some of those situations, though, the nature of the work may not be generated by the technology itself. That's fair. So there isn't as much maintenance and, you know, tending and feeding to the tech itself. But what does it change in terms of the factors of production, of human flourishing? So if you take for example, software engineering, right. Augmenting of coding, we're not going to need any engineers anymore. Far from it. I think what you start seeing is engineers. I haven't met one who's employed and sitting there going, you know what? I can see an end in sight on my backlog and I've only got six months worth of work in front of me. And once we get through that, I'm done. Every software team that I've ever worked with has more work than they possibly get done. Every doctor has more patients than he wishes he could see. Every plumber. How hard is it to try and find an electrician and a plumber and a builder? So I think that the nature of power tools coming to builders, diagnostic tools coming to medical specialists, better engineering tools coming to software engineers enables them to do more, enables them to do things better. Now, do they always do things more or better? And what things are they doing? That's also known the productivity paradox, where people get new capability, but it doesn't necessarily get pointed to the core processes that underpin them. I do think there's some questions to be asked about where we currently associate value in work and the industrial revolution and the information revolution. If this is the intelligence revolution, what does value look like? The exchange of goods and services, what does that look like? The exchange of expertise, what does that look like? Again, I think some of these have been proxies for value. And you think about people charge for their time, you go, why? Because time is a proxy. Like, I can't really quantify the effort. So I'll say an hourly rate. But increasingly people are going, actually, I'll actually do value building. Well, I'll tell you how much the whole thing will cost you and how much time it takes me is actually inconsequential. It's those kind of shifts that I think happen in the nature of work. But I don't see anyone, you know, the classic someone on the deathbed. No one ever said, I wished I worked more. Right. But there's also nobody who I know who's coming and going. I've got no work. Even in the situation where someone's unemployed, where someone is in a position where they're not currently in paid work, that becomes an interesting challenge from a social standpoint to say, okay, what is work? Is work employment or is work human flourishing and creating something and in the image of God? So I think there's again, big questions that is worth wrestling with universal basic income or universal high income. And some of those solutions I don't think solve the problem. I think they are just another trade off. And is it a trade off that we want to make or that's a conversation worth having as well? [00:37:07] Speaker B: To me it seems that there's something though we've lost our authentic human anthropology if we're even having that conversation. Like we don't grasp what a human person is, what enables their flourishing, Even just the fact that they can work and have meaningful no matter what that work is. But you know what I mean, to take that away and say, well, don't worry, we'll just give you money. [00:37:26] Speaker A: Yeah, it's kind of missing some of the points. [00:37:29] Speaker B: Yeah, the whole essence of what personhood is and what it is to find human flourishing. Already we're in trouble. We're having that as a conversation. [00:37:37] Speaker A: I mean, what is, what is to create versus to produce? Right. Because if you think about we are made in the image of the Creator and to create is an inherently human activity. We also infused with meaning. Infused, absolutely. And we also. And it's a divine act. [00:37:50] Speaker B: Yep. [00:37:50] Speaker A: Right. Producing or reproducing. And I don't mean reproducing. Producing something that has been created is a mechanical act. It's something which again humans have been able to do, but it's not the same. You know, the Pieta has been created once. I can create reproductions of it and I can optimize that mechanism as much as I like, but it's not the original art. You know, you're an incredible musician, you can write music, create music. I can listen to a recording of it and a reproduction of it. So I think we entangle, especially in work context, in a professional working context, we say productivity. Productivity, which is focused on producing. [00:38:28] Speaker B: Yeah. [00:38:29] Speaker A: What's the product, what's the output? [00:38:30] Speaker B: How many widgets in an hour? [00:38:31] Speaker A: Exactly. [00:38:32] Speaker B: Who's buying them? [00:38:32] Speaker A: The creative aspect of it is something which is come into modern work in the 21st century, probably a little bit more than it has in the past. [00:38:40] Speaker B: Well, ironically, I think it's In a sense, we're grappling with something and you see sort of glimmers of attempting to recover something which prior to the Industrial Revolution and sort of really the whole unbridled capitalist mentality approach to things, we would have had. You would have been more connected. The whole family would have been an economic unit together. They would have worked land that they were local to, you know, all that kind of stuff. And there was a connection. You said, I am responsible for the beet harvest this year, and that's what we do. And we plant, we sow, and then it was all integrated into your life. You weren't just simply, this is why I think socialism and capitalism, sorry, socialism and you know, and its more extreme forms like communism, why they haven't died. Because Marx was very good at critiquing what was a problem on the back of the Industrial Revolution. His solutions are terrible, but he can [00:39:35] Speaker A: point out the fact that's why still [00:39:36] Speaker B: today, his critiques often of capitalism and the system in general that we're sort of living under, they are actually quite prescient. His proposal is terrible. And so there's a much bigger conversation. We've sort of at times even like Covid, right, where people suddenly realized, ah, actually maybe it's not good to be at work the whole time and not with my family. But we've lost the bigger essence of what it was prior to that. [00:40:04] Speaker A: I mean, again, you look to, you know, the 1980s and the kind of, you know, capitalism's heyday and having this kind of takeoff and everything else, and it was, you know, greed is filtered. Yeah, yeah, exactly. Right, right. And filters into, you know, if I see your desk, you must be working kind of mentality and you can't leave till the partner leaves and all that kind of stuff. If as a byproduct of asking some of these bigger questions that our invention of the latest form of AI is forcing us to ask, we start to unwind some of these cultural holdovers of like, a worse time. Fantastic. Now, what do we come up with as a solution? To replace it is often the big challenge. [00:40:43] Speaker B: Right. That's the sacred transcendent thing that's missing, I think. And that's, for me, is my biggest fear. It's happening at a time when we have lost our sacred, transcendent sense of reality. We don't look to God ora et labora, you know, prayer and work. Yes, well, we've got plenty of work, the prayer component. And then when you have an artificial technology, mechanized and through robotics as well, with that, you know, it's that that prayer is just a fundamental. It's a discernment thing. Joseph Pieper, the Catholic philosopher, talked about. Prayer is like this profound and embodiment of hope even, you know, because it does. Just praying is a mechanism of hope in and of itself that's missing, you know, and then we're grappling with this big other big question of, well, what does labor look like? [00:41:28] Speaker A: So from an evangelical standpoint, that's where I think there's a golden opportunity. I think that these questions that people are starting to wrestle with, people who have never would have wrestled with the question in the past start to wrestle with, oh, I've got to ask these questions. Where do I turn for answers? And we're sitting here with the answers, the opportunity to get out there and say, well, you know what? No, you are far more than this material matter. You are more than just the product of your work. And, you know, you don't lose value in God's eyes because of how much you earn or don't earn. [00:41:57] Speaker B: You're not homo economicus, you're not economic utility. [00:42:00] Speaker A: Right. So there's. So we've got these answers, and I think people are hungering for it. We're seeing the trend of people returning to the faith or coming to the faith and coming because they're looking at the world and going, you know what? Increasingly, if. What was the stat I saw the other day, 52% of the content created last month was generative AI infused. So you're now getting to the point where the snake starts eating its tail. Where so much of the Internet, there's a thing called the dark theory of the Internet, or the dead Internet theory, where it's basically bots talking to bots, right? Yes. And now it's agents talking to agents. But, you know, there's all this kind of stuff where you go, is that real? Probably not. But is there a kernel of truth there? And as people say, I don't know if that's enough for me anymore. And we go, you know, the answer isn't. That isn't enough. But you know, who is, who will always be enough and who, you know what, who loves you no matter what, unconditionally, Unconditionally. In that context, you go, that is hope embodied and embodied on the cross. Absolutely. But actually embodied in people's lives. [00:42:54] Speaker B: Yeah, exactly. [00:42:55] Speaker A: So the opportunity for us to say, you know what, it's interesting, it's a bit scary, it's a bit unknown. But I'll tell you what is known is over here. That transcendence that you say is lost. Rediscovering that, re. Amplifying that, getting the message out to people, I think is critical. [00:43:10] Speaker B: Yeah. And I think the challenge is because you've got to decouple those, well, from the machine, in a sense. And that's the hard bit. [00:43:16] Speaker A: Sure. [00:43:16] Speaker B: As we're seeing. And we all know that social media, before we even get to AI Social media has us short and curlies, and it's hard to let go. [00:43:24] Speaker A: It is, it is. But I mean, even I think, you know, the church helps with things like that. So, for example, you know, Lent for me is always, okay, I'm going to give up social media. And inevitably, you find there's a reason why all those, you know, 12 steps and other programs like that that help people break addictions while they do actually work. If you get that distance, you start to say, well, hang on, maybe it is a little bit different here. Being even more mindful and purposeful with the use of the technology itself. It doesn't need to be all roads leading to one black mirror. You know, the fact that if I'm reading my Kindle, I'm obviously reading a book. You know, if I'm using an iPad and I'm drawing on it, I'm obviously writing. You know, there's these signals that we can send to the people around us, but also to ourselves. This is what I use this for, and this is what I use this for. [00:44:05] Speaker B: Yeah. Is there a. I sense there's a point, though, at which, because of the power. And this is the thing I'm very aware of, like, C.S. lewis talked about science and really the technology it produces as the magician's twin. [00:44:18] Speaker A: Yes. [00:44:18] Speaker B: And he said, look, the magician, they don't really have power. It's an illusion. They don't make our rabbit disappear. It's all fake. But a scientist can give you real power, and that power is alluring, and it is corrupting if you get too much of it. And, you know, the danger with the magician's twin is that that there's real power produced and people want to wield that power like so. We were talking earlier about democratization. The Internet, it seems to me, is arguably. I think Dan probably pretty much has been a democratized technology. Anybody could sort of play in the sandbox, so to speak. Within reason. This doesn't feel like that, though. To me, though, it feels like you've got a much smaller group of much, much smaller group of people who actually already have the keys to the car and they're not going to let anybody else drive it. In a sense. [00:45:07] Speaker A: It's an interesting question because if you think about with the models that we talk about, there are obviously certain names, certain brands that people understand and kind of gravitate towards because they have been involved at the very opening up of this next stage of AI development. But there are just as many if not more open source models. So there are democratized access to capabilities, which is for free, roll it yourself, play it on your own machine. You don't need to be using the data centers. Is it as capable, is it at the top level of these other models? Not always. But is that what you need for the job you're trying to do? Not always. So I do think there is a different. I'm not trying to diminish the fact that there is a locus or like a pull together in certain capabilities, particularly for training some of the models as well as some of the inference which is the using of the models where you disproportionately skew towards certain platforms or brands. But I don't think that's the end of the story. I think that with a lot of these things, the early days of the Internet, it wasn't democratized at all. It was entirely academic and military. And then what it became. And when we say the Internet, what we really mean is the web. And so that's the same thing. When you think about this, we go with AI. Well, what we really mean is large language more. [00:46:17] Speaker B: Does that mean though that it seems there's a hook in this potentially. [00:46:22] Speaker A: Go on. [00:46:23] Speaker B: That you now also, if you are a bigger player, you now wield a technology that can actually be used against those who want to have a more democratized version of the technology. You know what I mean? Like the technology itself can be utilized to target the vulnerability of an open source model, could it not? You know, to basically then you get power back that way in a sense. [00:46:44] Speaker A: Not sure if you like. If you think about the nature of how a large language model works, where you're trying to get the weightings, which is the statistical numbers that are telling you what token comes next after this one. Those models, once they're trained, that isn't something that can really be constrained. And this is what you're seeing in some parts of the world which they don't have access to for whatever tariffs or reason, they don't have access to some of the other models, they can still access the weightings of these models and still use them. So I don't Know, if you can quite use one to block out the other. I do think back to sleepwalking. Will people always go and do the extra work required to get access to them, or are they just saying what's being provided to me and right in front of me and I'll just use what I've got? That's been a problem well before AI, so like, you know, all the examples around, people would give away all their date of birth and their first name. Anything else for a Mars bar coming out of the London Underground. And so you go, there's the, you know, let's just say we're an efficient species when we want to be in terms of reducing the amount of effort required to achieve an outcome. And so I think in the context of AI, I think that there are certain technologies which actually what certain aspects about what it's doing is unlocking huge democratization. When ChatGPT first launched, it was basically a tech demo. Like they sort of said, hey, look, we've got this thing and it looks what happened and it's trained. And then the world lost its mind for the next couple of days because like, have you seen this thing? Have you tried this thing? And fastest ever to get to a billion users worldwide. But the nature of it wasn't designed for that. It was just that we ran with what we saw and what we were able to uncover. I think that taking that protein folding example I said before just releasing all known proteins. Here you go, we'll just run them all. Here you go. Which is what the Nobel Prize was partly awarded for. That's huge. So suddenly you've got universities and medical research facilities around the world having access to capability that they never had access to before. So I think it's not a black and white. Is it democratized or not? I think it democratizes different aspects at a different rate. [00:48:35] Speaker B: Because the thing with it too, is that the potential that you have to wield control a very small group of people over a large group of people with this technology, I think is. It feels unique to me. [00:48:46] Speaker A: Okay, why is that? [00:48:47] Speaker B: The scalability of it, the speed of scalability. You know, you have a. Elon Musk can own his own engine effectively, that can not just propagandize, but actually drive behaviors and all sorts of stuff very powerfully, it seems to me. And the potential for what that means for just basic human equality. [00:49:15] Speaker A: See, I see it slightly differently. I think that there are definitely people who, as with any technology, any information technology, there are usually a smaller group of people that are creating or that are forming the tech itself. But then as it spreads, it usually evolves and develops and extends and changes. So it's very difficult for any one player, I think, to hold the entire model together. Which is part of the challenge with some of these capabilities as well, is that if you did want to constrain them, if you did want to say you can't have access, it's very hard to do that because kind of the rabbit's out of the hat, the toothpaste out of the tube. [00:49:50] Speaker B: Could you make an argument though, he who controls the land and the data centers effectively could wield that power because they could just say, look, no more data centers. And then there's only so much computing that can be done. Do you know what I mean? [00:50:02] Speaker A: So, yes, but depending on again, the type of computing that's needing to happen. So a lot of the, if you take for example, the Tesla example, a lot of the power of having the distributed computers. So the fact that all the Teslas themselves have computers in the Tesla, the distributing computing can actually become in itself quite powerful. I think that the fact that we have computing at the edge or computing that we have access to means that if we were just dependent on it. So go back early in my career working in banking and a lot of the time there was mainframe architecture. So everything you had was a dumb terminal and what it was doing was connecting to the mainframe. And if the mainframe shut down like it does in all the movies, then suddenly everyone can't use anything. But then we don't use mainframes. Still very much are in the picture. But most of what we do is localized computing. It's on the device. And this is something that actually Apple, I was reading a piece where it was saying, have they stumbled into one of the smartest AI strategies by being late to the game? Because they've been able to see what's going on and what they're trying to do is actually pick the right lane, pick the right lane and also constrain it. We want to use it here and here, but not here. We want to have the private compute, etc. So I think that we'll see how it pans out. But I think there's enough drive for people to want to be able to have local control, open source. I want to see it myself. I don't want to trust this without me being able to see everything in it. That movement hasn't changed with this version of AI, the access to the data and the compute to train the models, that is Something which is not something that everyone can do, but training the model and saying we've got this really great model, but then no one can use it. You spend a lot of money for something that you can't get any value from. How will that play out? [00:51:39] Speaker B: It seems to me too there is a question about how much governments are already ahead of the curve in a sense. I mean, governments are notoriously slow often. It's often private industry that's a bit more efficient, right? But this technology, because of the power that you can wield with it, I know there's already. I don't know whether it's gone beyond the stage of speculation, despite what some are claiming, but that some countries are using AI, for example, for double tap missile attacks and things, you know, so it's not even a human who's. They might target. Create a target packet with human intelligence and then the fact that they hit and then like 20 minutes later, whatever is the AI is the one that's doing the second tag or whatever you can see like even the potential. And we saw this during COVID the ability or the desire to corral populations. I mean, imagine a lockdown with an AI technology that's really at the surface of, of those who want to really keep you locked down. You know, you can see a nefarious purpose. And I just, I wonder, I do wonder. Human power, the corrupt nature, you know, C.S. lewis again, the version's twin. [00:52:43] Speaker A: I do wonder inside of that. So while I'm not familiar with that particular example, I think that to quote the fantastic philosopher Dr. Ian Malcolm of Jurassic park painting, sometimes we're so worried about whether or not we could, we didn't stop to ask whether or not we should. And so I think that that dilemma is very much on the right in front of us, right? The nature of what these things can do versus what they should do is at the core of kind of what we're talking about. I wonder, to the point around the power that we're talking about, how much of that is connected to things that none of us have control over. Like I can't launch a nuclear weapon, thank God. I don't think anyone should be able to, but I definitely can't. So there is a very small group of people with pure executive power who have the ability to unleash destruction that the world doesn't know, right? We say and we go, okay, what's stopping that person from doing that today? And we go, oh, the law and the safeguards and things. I go, actually, I think it's Human nature that ultimately is stopping. Do you want to be the person who did that now? I love the movie Crimson Tide. You know, like, you know, at a certain point you go, what happens if the safeguards fail? What happens if someone makes the wrong call from the wrong signal? What if, what if, what if the same challenges apply with any technology that we deploy and if we're putting it into positions where it could be cause a detrimental outcome to humanity, We've got to ask that could should piece the control, the corralling of populations control. That's why I think it's important to be able to create conversations. You know, the kids these days say go out and touch grass. It's conversations to go out and actually connect with people in the real space. Like, I am very sure that you are here. Yeah, I am very sure that you are not an AI. It's an embodied experience now. It's messy and it's organic and it was all these kind of. It's amazing that rediscovery, which I do. I see plenty of signals. You know, part of my job, looking to the future, I see signals that re emerging analog bags where people are like, I don't want to be on my phone all the time. In London, they were doing a study where they gave people just a piece of glass that's the same shape as a phone. So they could do the scrolling because actually it's become like nicotine patch to break out of people having an analog bag with things that they do when they're not on their phone. A book, knitting, cross stitching, people putting CD players back into their cars. You see these signals emerging of people saying, actually I don't know if I want to choose that future. So in that same context, writ large with some of these bigger choices, I don't want to be influenced just by what I read in my social media bubble. I will challenge myself to go find the opposite view to that I love. I was talking recently with our team when I said Saint Augustine. Saint Augustine or Saint Aquinas. Saint Thomas Aquinas. Saint Thomas Aquinas would always would only state his argument after he could state his opponent's argument better than they could. That respect for the opposite position is not something that you see in our daily discourse. [00:55:23] Speaker B: Oh no. [00:55:24] Speaker A: So when someone comes and says I've seen this stuff on X or Twitter or Blue sky or Instagram or whatever else and I'm like, cool, have you touched some grass too? Like, have you kind of spoken to some other people I see with my own kids, like My teenagers, you know, they'll come say, oh, I saw this thing, and I'm like, great. What's the strongest argument against that? And then how do you sit in the tension of those two things? That's critical thinking. How do we train that to make sure we don't lose that? Which would be sleepwalking. [00:55:48] Speaker B: How does that. So I like the sleepwalking analogy. I think it's very apt. How it seems to me, though, there is probably going to. My suspicion is that there will be larger groups of people who will just say, we are too. The technocracy has us too much and we want to just pull right back, you know, so more than just an analog bag, but let's go and buy land and farm together and all those kinds of things. I sense there will be. There's probably a coming moment where there will be a fix, effectively a quiet rebellion, you know, the quiet checking out of technology. [00:56:21] Speaker A: So there's again, if I think about signals of possible futures, there's definitely signals where you see people taking digital fasts and they don't come back from that fast. When they start to see their life as actually fine without their digital technology. You see the rise of the sale of dumb phones, right? People wanting to break that technology grip on them, old school books. And at the same time, what's happening, I think you see signals the opposite direction, where the digital world is becoming more artificial. So Amazon now restricts how many novels you can submit in a day. Because what's happening is people are generating them and just slop. People are starting to go, actually, I can't tell whether that influencer is actually human or not. And I was talking with a friend the other day and they were saying about how angry they were because they'd seen a bunch of ads in a magazine like, like a printed magazine for a women's skincare product. And then when you read the fine print, it actually says the model's generated. So the model doesn't have pores. So no wonder you can't see the pause. It's not the product, it's the actual subject. So I think there's the sense of people kind of going, okay, hang on, wait, what? Wait, what? And I hope that what we see is by unlocking additional productivity and capability in the workforce. Because people would argue I'm just too busy to think, but they wouldn't probably say it that way, but I'm too busy to think about the big questions. What if you're not as busy on [00:57:39] Speaker B: it, just trying to survive, just trying to earn money. [00:57:41] Speaker A: Just trying to get through. Just trying to get through. Well, how much of that is you trying to prepare for the meeting that actually doesn't need to take place because now we've got a different way for that process to operate? How much of that is us saying, okay, you know what? You're not taking minutes in the meeting anymore. You're thinking about what the meeting's for because something else is taking the minutes. What does that do? What does that unlock? If you take those moments where you go for a walk without your AirPods in, without your podcasts, even a great podcast, what happens where you learn about being bored, you learn about. You know what I've never noticed how beautiful leaves are. You know, all this kind of contemplative [00:58:10] Speaker B: creativity comes back because you need boredom for that. [00:58:13] Speaker A: And it comes back faster than I think we give ourselves. I think some of the fear is maybe a little bit overstated. [00:58:18] Speaker B: Do you worry, though, about the potential for. I'm not even talking about any sort of super intelligence here. I'm skeptical that sort of thing is even possible because of what a human person is and what intelligence actually is. [00:58:30] Speaker A: Is. [00:58:31] Speaker B: But I do recognize the power to delude people at scale because of the power of the technology. Do you worry about. I know some people have the conversation about the possibility. If you wanted a perfect Antichrist type delusion in the world, this would be the technology that could deliver it. And it would just have to create a sense of magnanimity and offering people true salvation in a sense that, you know, a utopian, which would be dystopian. Do you. Do you ever does that? I mean, it's obviously a possibility, but could you ever see it? You know, people being organized enough to pull that sort of stuff. [00:59:05] Speaker A: Do you know, I find we, you know, we're talking before about conspiracy theories, and one of the reasons why I don't think any of them has ever quite taken hold for me is the fact that I have been around long enough and worked with people enough to know that the ability for a large number of people to hold, to hold to an idea, an organization, I mean, even the coming of one of the things that I remember studying. When you study Scripture, one of the things that gives historical weight and veracity to the Gospels is the fact that all four of them don't get their story entirely lined up every time, and [00:59:37] Speaker B: they contain details that are embarrassing to the authors. [00:59:39] Speaker A: Right. So there's that too. There's that too. And so you have the sense of, like, there are these little Glimpses of what I see in the human condition, or maybe it's in society or it's how we're structured that make me go, oh, maybe we're not going to be as easy to fool as I might think at first. Wake now, if enough noise is being made to wake people up so they're not sleepwalking to my fear before. [01:00:00] Speaker B: Yeah, and that's the key. Your sleepwalking analogy is so apt because if you are already in the stupor. [01:00:06] Speaker A: Yeah, so. So I do think that if we weren't having clanging cymbals and, you know, ringing gongs and waking, trying things to try and wake people up, alarms and everything else to wake people up, and we were all just like, you know, the guys from Wally on the boat just consume, consume, consume. But I already see signals that that's not the case. Like, we are already at a point where if we wanted to have that, we could. You can have anything delivered to your door using a magic glass mirror. You know, we've got that right now and people are going, I don't know if that's what I want. And so the fact that we're asking that before we get to even more, even more, even more, that gives me hope. That gives me hope. And I do think that there's an element of we are in a world which is broken, which is challenged, but at the same time is beautiful. You know, we are ourselves, people that can always do better. Yeah. But that doesn't mean that we don't have a proclivity to do things that aren't good. And so I think you look at that and you say the wrestling with it is just like the development of us as, you know, when we're babies growing into adults to eventually being with God directly. You go, okay, I think there's hope there. [01:01:10] Speaker B: What are the implications for the church? [01:01:13] Speaker A: A lot. I think we talked before about, I think the mission, the evangelical mission is really important because we have the message and the message can actually help a lot of hope. I think that a lot of times I've seen, particularly inside of the Catholic Church, the challenge has been having the right kinds of resources. So outside the church, people would look at and go, oh, the church has all this money and all this land. And I'm like, well, no, the church has a lot of land. But that is not the same as being able to stuff that's been given to it over years is not the same as being able to leverage that into expenditure to go help where all the help that's Needed now, you know, you've got organizations like Caritas that are amazing and do incredible things. Could they do more? Could they reach more? You think about the nature of digital ministry, something that you and I have spoken about for many, many decades at this point. [01:01:56] Speaker B: Narratives. [01:01:57] Speaker A: Yeah. And I think about how hard it has been in the past to try and get the church to build, you know, I don't know, digital properties or websites. How much does that start to. Does the calculus shift? So you go, that's really interesting. The other side, though, is where do people fill the God gap with this? Your point on narcissism is really valid. The other one I was using, to use a Harry Potter metaphor, which my kids are very proud of, because I haven't actually read the books, but I have seen the movies. The mirror of Iris said, which is desire backwards. And from the movies, if anyone hasn't seen them, you know, that point around, they stare into the mirror. What do they see? They see what their heart desires the most. And they can. They can waste away staring in front of the mirror. So the fear for people. For me, the sleepwalking metaphor is people don't wake up and they're like, I'm actually happy being asleep. I'm good. And you go hypnotized by it. Hypnotized. So the church saying, like, I saw examples, and they're usually just done for clickbait headlines. But they're examples where the. You know, actually there was the one the other, other week, a Buddhist monk that's a robot. And the point was that wasn't for clickbait. That was him saying, well, they can't get more people becoming Buddhist monks. And so they need people to still go. They need a thing to go through the mechanisms of it. Right away you get into this tension of, wait a minute, going through the motions versus, you know, if I have someone that can make the sign of the cross, does that mean they're praying? [01:03:13] Speaker B: No. [01:03:13] Speaker A: Exactly. So I think the church not embracing this for the best possible use. It's like saying we're going to use a word processor, but we're not going to use spellcheck. But that would be silly. Like, we're going to use these capabilities as tools to unlock the next level of human flourishing. Fantastic. We're going to ignore them at our peril. We're going to allow others to put this technology, this tool in place of the church. St. Augustine was saying with his. Now my pronunciation is awful, but he talked about, is it ooty and free, though the difference between the utilization and the ultimate free fulfillment or the joy. And his point was the tool, using the tool is good. [01:03:53] Speaker B: Yeah. [01:03:53] Speaker A: Having joy in the tool. No, the joy belongs to God only. [01:03:57] Speaker B: Yeah, that's right. [01:03:57] Speaker A: And so it's the sense of, like, don't fall in love with your tool for the tool's sake. No, it's what the tool enables you to do. [01:04:02] Speaker B: Yeah. And we share in that. Aquinas would talk about that idea of we share in the very life, the very beatitude of God. That's, you know, that's profound. [01:04:09] Speaker A: So I think the implication and what I'm really excited about, I have to say, from a church standpoint, I talked before about, briefly about Father Benanti and the rome call to AI Ethics, which was very early on in this, before ChatGPT, the Minerva Dialogues. Like, you're seeing the Church, Pope Leo himself coming and everyone's waiting for it. [01:04:26] Speaker B: By the time we publish this, it might well be out, actually by the. [01:04:28] Speaker A: Well, Antica Novo, which I know you've spoken about before as well. Fantastic, fantastic positioning. Pope Leo picking the name Leo. So, like, I do not have a fear that the Church Magisterium is ignoring this at all. In fact, it's almost like this is what we've been gearing up for, for hundreds of years now. The conversation is there to bring that to the fore. I think the everyday Christian, their job is to look at this and go, you know what? You can have conversations with colleagues, family and friends to say, there is a bigger reality here that no release of any version of any model is going to take away because it's just not possible. You cannot engineer a soul. [01:05:05] Speaker B: No. And that's. I think the interesting thing too is that there will be. There's obviously clear boundaries. We have. The sacramental economy can never be interfered with with artificial things like that. You'll never have a confession bot or any of that kind of stuff. One thing I have to say is I did see a church and I actually, I think it was a parish, actually. They had like an AI bot. [01:05:24] Speaker A: Digital Jesus. [01:05:25] Speaker B: No. On the front page of their website, it was. And people, I think they were right to say, hey, what are you doing here? You go to your parish website and you're engaging with this AI bot thing. [01:05:33] Speaker A: What was Catholic Answers had an early thing where they had Father Justice. [01:05:37] Speaker B: That's right. [01:05:37] Speaker A: So Father Justin, it did not go down well. So then what they redid with. And now it's really. Even just three years later, they did a retake on it. So he's not. Well, they did it far sooner than that, if you look at it. Now go to Catholic Answers. I like Catholic Answers. Don't get me wrong. It's now like layperson Justin, but it's not even that. It's like. It's an interactivity of bringing that data together. Magisterium AI is a good example, bringing together the teaching of the Church and the same stuff. Do you remember years ago, I was talking to you about the idea of the Damascus project, and I was like, how can you turn people on 140 characters? How could you turn someone's position on a particular matter of the Church? And I remember you saying to me, mate, you've seen many wise things to be over here. But I remember you saying to me going, you can't summarize the whole beauty of the Magisterium, the church in 140 characters. And I said, I agree, but could you get someone to then go a little bit further? [01:06:23] Speaker B: Could you spark? [01:06:24] Speaker A: Could you spark? Could you get them to go, okay? You know what I mean? I think we were talking about the time it was the vessels used for the precious blood. And should it be glass or should it be some of this? And it's like, well, I can quote you this Magisterium, the precious blood is precious. Put it in something precious right now. Does that summarize the teaching? No. It starts. You go, whoa, let me think, let me ask more questions. [01:06:47] Speaker B: It's the heart of it, though. It's the. What you're describing there in that brief statement is the impetus, which is love. [01:06:53] Speaker A: Yes. [01:06:54] Speaker B: Love for God. [01:06:54] Speaker A: Yes. [01:06:55] Speaker B: And it's a desire to move closer towards the good and what is true. [01:07:00] Speaker A: Yeah. [01:07:00] Speaker B: And that's. That has been summed up so in [01:07:02] Speaker A: the situation where I can use these tools to take thousands, literally plural, thousands of years of teaching and summarize it in a way that makes me start to understand the essence that is. That is good. That is good. And, you know, I remember saying, when I'm being asked in the past, why am I. Why am I still Christian? Why am I still Catholic? You know, as I get older. And my response was, every time I've asked the Church a question, really asked the question, I've never come back with yes, full stop or no full stop, it's always been, here's a tome of writing. Here's saints, here's holy people who have spent their entire life contemplating this. Here is why we think the following, and here is the principles behind it. And it's like, I can't argue with that logic. So I think there's golden opportunity for the church that's presented with these tools and capabilities. There are challenges, sure, but I think that with any of these trade offs, if the focus becomes on keeping eyes on where they need to be on God, I don't see how it goes bad. [01:07:57] Speaker B: Yeah. And I think there's a. One of the things that I think is going to be kind of interesting is like you talk about the whole idea of people sleepwalking and there are more people who are starting to wake up, I think. And I think there's a real opportunity now for the fullness of Catholic social teaching, which I think quite. It is kind of embarrassing at times in some court has been reduced to just questions of are you actually providing for the poor? So the preferential options for the poor and then environmental issues of late, like that's part of a much bigger whole which is grounded in the human person, the dignity of the person, the family, the fundamental first community that I think, because this is the only way you really grapple with this technology and push back against it as you actually recover authentic community which has to begin with the family, a sense of human anthropology. That's exactly what lies at the heart of that teaching. I sense there's an opportunity for. In much the same way Ririm Navarrem, you know, Pope Leo XIII really does awaken to the whole dignity of work in the human person. You know, there's something similar here, I think. [01:09:02] Speaker A: I think absolutely. And I mean, I remember talking, you know, Pope Benedict said, you know, the harvest is plenty, the workers are few, and the challenge was the new springtime in the church. And so if you take that metaphor a little bit further and you say, well, it's a lot of. You can harvest a lot more when you've got, you know, good productive land that's fertilised the right way and you've got the right kind of machinery to assist, you can get more, more from the land. When you say the work, when we say the workers are few. Yes, but those workers can do a lot more. For a while I've had this idea, I've talked to a few people about the loaves and the fishes and so this idea of the baskets. 21, 20. How many baskets was it? 14 baskets. I forget the number now. 12. 12. 12 baskets. 12. Thank you, mate. 12. Magic number. Holy number. I just checked my. Yeah, yeah. [01:09:43] Speaker B: 12 baskets. [01:09:43] Speaker A: 12 baskets. So, hey, all I've got, all I have, Lord, is just this, a couple of fish and some bits of bread. That's all I've got. You've got 5,000 people. I can't be any help. And at the end, there were 12 baskets left over. Like, God never takes us, whereas, you know, grace can't keep us, is the phrase. But he never takes our offering and goes, is that all you can give me? [01:10:04] Speaker B: Yeah. [01:10:04] Speaker A: Right, so if we can now turn around and say, I can offer this to you, Lord, to you, I'm offering this. Do with it what you will, and leave it to him, not to us, then we're not relying on human proclivities and other challenges. We're turning it to the divine. So I think that presents massive opportunity. [01:10:20] Speaker B: Yeah. It's interesting because there's a whole, like. I know I saw someone recently, a mutual friend of ours, actually, and he. He was very. He's a philosopher. And his question was like, even the question of engaging with AI, you're conversing, you're not really conversing, you're communicating of sorts. [01:10:36] Speaker A: Yes. [01:10:37] Speaker B: But he's like, well, it's. It's not a being. There's a privation of being, the absence of being. In the Augustinian line of thought, that means the absence of. Of good. [01:10:45] Speaker A: Good, yeah, Is evil. Yeah. [01:10:47] Speaker B: You know, and he said he's not. He didn't come down and land on a position. But he's like. It raises an interesting question for him about whether even just our dialoguing with it is, in a sense, is not a healthy thing. [01:10:58] Speaker A: So interesting. I'm. I think I know who you're talking, but I would. I would go down the path and say, let's rewind pre the ChatGPT era. [01:11:07] Speaker B: Right. [01:11:07] Speaker A: And I code up a website with a bunch of fields that you have to fill in. Now you're filling in those fields. You're not conversing with me to tell you those fields, but you are putting words exactly right. If you're having to fill in a document that you're having to outline for applying for a job, you're like, well, I'm actually conversing, quote, unquote, with a machine. The machine's not conversing. I've had plenty of conversations where people haven't been talking the other way back. And so maybe it's more a monologue than a dialogue, but it's because of, I think, the mimicry. It's because of the sense of, wow, this thing feels like it's responding back to me. [01:11:39] Speaker B: Seems to be right. [01:11:40] Speaker A: And it's not Clippy. And it's all these kind of, you know, Those rudimentary models that we had. [01:11:44] Speaker B: Yeah. [01:11:45] Speaker A: This feels more sophisticated. You've talked before about the uncanny Valley, and it's true, still sits there, particularly with some of the visual stuff. But from a text standpoint, the reason why people say the Turing Test is done is because you go, well, I can't really distinguish between the two now, ironically, we now can, because we're going, well, it's using the perfect option, Oxford comma and too many EM dashes, you know, and so, like, it must be right. Which is annoying for me because I like the M dash, but there's this sense of us having to. What I said before, reprogram the shortcuts. [01:12:13] Speaker B: Yeah. [01:12:14] Speaker A: Are we. If we started, say, I'm talking to my friend. And, you know, you've mentioned again on your podcast before about a lot of our prior art on AI comes from the movies and the, you know, books and things. [01:12:26] Speaker B: Yeah. [01:12:27] Speaker A: Everyone wants Jarvis. Let's just park for a moment that Jarvis and Ultron become a thing and everyone wants Jarvis. In that context, you go, well, actually is Jarvis. And in the movies, yeah. Jarvis becomes embodied. Right. There's the sense of us saying, well, I've got my AI, I've given it a name and I've given it Persona, and I've given it these ideas. And you go, well, you haven't. What you've actually done is you've provided prompting and guidelines and guardrails to constrain a statistical model to only weight more towards these and not these. That's what you've actually done. It's not magic, it's just maths. But Arthur C. Clarke, any sufficiently advanced technology indistinguishable from magic. You talked before about the magician's twin. [01:13:07] Speaker B: Yeah, Louis. [01:13:08] Speaker A: So I think a lot of this challenge inside of it is our own perception of what we think intelligence is. [01:13:14] Speaker B: Yeah. And that's that sleepwalking analogy. Why it's so apt. It feels to me like we are a little bit in the valley of dry bones. And you're like, come on, wake up. Sleeping is because. Yeah, it's clear that for some people, the stupor is already upon them. And so AI has just. Sadly, some cases ended very tragically. [01:13:30] Speaker A: I absolutely agree. And not to diminish that at all, because I think that those warning bells are ringing loud. But I actually, if I was being asked on a couch, tell you what I think. If I zoom out, I actually think that this moment in AI may be the wake up call that we needed to get out of the social media stupor. Out of the digital zombie stupor, out of the mobile device fixation stupor, that nothing was breaking, that we continued to go further and further and further into a more and more artificial world until we eventually created an artificial world that was so rich with artificialness that we go, I don't know if I like that. So I actually hope for zoom out. This is the reset that's required. You know what? I'm going to let my tools go work with the other tools to get the outcome I need. And I'm going to focus on the things the tools can't do, because what they can't do. I've spoken about this before. We talk about an AI, about having a human in the loop. That's the context, that's the word, that's the phrase that's used in AI governance. Having a human in the loop inside the process to do the final check. Fantastic. But sometimes we're talking about having a human at the heart. It's not a place that's the domain of anything artificial. This connection between two friends have known each other for years. This is not the domain of something artificial. So humans at the heart of this. There are examples where people are sending their AI to meet their dates AI and the two AI have a date and come back and report whether the human should meet. And we go, well, that's kind of ridiculous. But if you apply for a job and you use AI to write a cover letter and you send that into the job and they use an AI to assess the COVID letter or even [01:15:04] Speaker B: to zoom you if you get through, [01:15:05] Speaker A: to tell you whether or not you should have the zoom, it's the same thing. And so this is, this could should debate that. We are going to have to wrestle with time and time and time and time again. [01:15:13] Speaker B: And the technology, the science, can't tell you what ought to be. It tells you what it is. It can give you power, can't tell you how you should wield it. [01:15:19] Speaker A: And when we're talking about humans at the heart, that is that guardrail that says this is not the domain of the artificial. And that I think we are as a species examining that a little bit closer than we ever have before. [01:15:30] Speaker B: And young people in particular. There was a survey conducted about a week and a half ago in the states, just over 3,000 young people. And what was interesting was, and it was all down to the technological, among other things, technology was a big factor. They said they would much prefer to live at a time in the past. [01:15:44] Speaker A: The majority did, yeah. I mean, there's Headlines, you know, why does Gen Z hate AI? And you know that the adoption is actually coming from boomers more than it's coming from Gen Z or Gen Alpha. That these are the generations who have had a far larger portion of their lives. We're in that Covid artificial environment, and so they're coming out going, why would we ever want to have something that's like that? In addition to they're looking at their parents or their older siblings who might be the generation head and going, I don't know if I like all of this. In the same way that we were talking about how Gen Z and millennials were disproportionately representing anxiety over depression as being like the number one mental illness in the US a few years ago. And you go, wow, how have we got ourselves into that position where as a generation, you have more than any generation prior to you. All the things which Marx and all the others would talk about. If we just changed everything, then everyone could have everything comparatively to where they were at that time. People have. And yet they feel lost, they feel emptiness. Why is that? Well, because on the surface, you've got all your needs met, but your needs aren't met. The real needs aren't just your material ones. [01:16:48] Speaker B: Yeah. The fundamental need of the human person, obviously, primarily to love the. To be loved, to know God. [01:16:56] Speaker A: Yeah. [01:16:57] Speaker B: To give yourself to something greater than yourself. [01:16:58] Speaker A: And I think the nature of those things, those of us who have been on our own journeys of faith, as winding as those can be, have been wrestling with those, more so than some of our, you know, brethren who haven't as a species. We're now having to ask that question. And there are different answers. Different faiths have different answers. That's fine. But the common thread is that we're all going, yeah, there is something more. There is something more. And you either believe in, you know, the comedian was saying, you know, you either believe in nothing, which is basically everything. It's a pretty special. Special nothing. It's a special nothing. Or you believe in something. Yeah. And I think that in a situation where you've now got this model where every day, every day you are getting reminders, you should really be considering, is there something more to it than this? Because if there isn't, this is all very artificial. Useful, productive, sure, but artificial. [01:17:45] Speaker B: Yeah, yeah. Efficiency is not the same thing as flourishing. [01:17:48] Speaker A: Yeah, fantastic. [01:17:49] Speaker B: Not the same thing as flourishing. [01:17:51] Speaker A: Fantastic. In a situation where in a business, you're trying to achieve an outcome, that outcome is something. That's a process. That process can be automated. Brilliant. Let's go. Let's unlock more because we've got more work to do, more clients to serve, more people to help. Love it. In the context of me and my relationships, my connections with my family and my friends, I don't need a lot of help with that. Thanks. Maybe help me generate some stock photography for my invite to my birthday party. But don't be at the birthday party. [01:18:15] Speaker B: Yeah, yeah, yeah, yeah. That's a good way of putting it. To conclude, then, what would you say to someone, what's a piece of advice do you think is valuable, like, for your layperson? We're not talking here about the tech gurus who are going to be building this, the architects. You know, that's a whole other question. How do you not become a sleepwalker, then? [01:18:39] Speaker A: It's a great question. I think training yourself to be comfortable with the discomfort of not knowing is the beginning to challenging whether or not this is the future that we want to have. So when you are reading something that you disagree with or that you agree with, seeking out something, that is the challenging the opposition and then sitting in the discomfort of that is the thing that helps, I think, protect against just taking what you're seeing as red or taking what you're being presented with as being the truth. If we are looking at sleepwalking as being the problem, then the answer has to be to wake up. And so when you're hearing alarm bells ringing, when you're hearing the dogs barking and there's something in the driveway, don't ignore those signs. [01:19:30] Speaker B: Ignore it at your peril. [01:19:31] Speaker A: If you are worried about these things, have the conversation. If you are concerned about where these things might go, have the conversation. Who do you have that conversation with? If you've got people in your friends or family or network who are more technically aligned, you can ask them technical questions, but you can have conversations with your politicians and leaders, you can have conversations with your priests and pastors. You have conversations with your friends and go, you know what? I don't know where we're going with this. I'm not quite sure whether I'm still in a dream state or not. What do you think? And I think opening up to that conversation, that is probably the key thing. Just not accepting blindly what's being offered is the key, I think, to waking up. [01:20:10] Speaker B: You know, what I think is another key factor is that in discernment is actually, like you already said it, you get out and touch grass. [01:20:16] Speaker A: Totally. [01:20:17] Speaker B: If you are doing real things in the real world with real people that is a powerful discernment in yardstick because the artificiality of it becomes very apparent. [01:20:25] Speaker A: Absolutely. And I think being able to use the tools and constrain them to the domains where they can add value, you and then recognize that actually coming in the other way, not so much. So if you are finding the temptation to use these tools to write a 21st speech or something like something like that, maybe ask yourself, is this something where you're just the human in the loop, but actually the tool is the thing that's producing it? Or is this something which should have the human at the heart, in which case you write the speech and it may not be as erudite and it may not be as witty, but it will be authentic and it will be human. And guarantee the person would want that more than they would want the artifice. [01:21:05] Speaker B: Oh, it'd be better, I guarantee. I just. I think in every meaningful way it will be better. [01:21:09] Speaker A: I think what you opened with is very, very important. These, any of these tools. Look back in time, look over the last 20 years, last 50 years. Anytime we've come up with a new intermorifica, these miraculous things, as John Paul II called them, anytime we've come up with all of these things, inevitably we pull to utopia or dystopia. And yet every single time it's been somewhere in the messy middle. So I think the same situation appeals here not to diminish the fact that some of the characteristics, some of the capabilities, Arthur C K level, you know, they look like they could be magic [01:21:42] Speaker B: power is pretty impressive. [01:21:43] Speaker A: So you have to really want to wake up. Yeah, I'm not getting into a matrix thing here, but you have to want to wake up. I think in that context, it's a matter of saying what you're going to hear that's utopic or that's dystopic is probably not going to be the way it lands up. Or you have agency in deciding whether it will end up that way for you and your family and the people around you. [01:22:02] Speaker B: Well, ironically too, people always talk about 1984, 1984, 1984 and 1984 is an imposed dystopia. Reality is consistently what we've fallen into is brave new world, which is a self imposed one. [01:22:15] Speaker A: You know, sleepwalking. Yeah, it's the thing where we choose this. And I think to your point, even before, about, you know, the Antichrist situation, it's like, what are we choosing to believe? You know, when, you know, C.S. lewis and the end of the Narnia Chronicles, do You honestly think that's Aslan? Do you really think that's a lion? Or do you think it's the donkey? Like, what do you really think? And it's this emperor's new clothes comes back time and time again because people are willing willingly suspending their disbelief. And you stop that for a moment and go, well, hang on, hang on, maybe I shouldn't be doing that. And now's the time for me to start to retrain myself. And hey, when you start working out at the gym, you don't go straight to the heaviest weights. Start with the small stuff, start with the thing where you go, I'll use my brain first and I'll use these AI empowered tools to help extend that, not the other way around. Right. I saw there was a study that was coming from MIT that talked about that, that if you use AI to start you with the brainstorming and then refine it, you actually the students that they were testing on record less and their writing was worse. If they started first and then used AI to help refine, to research, to extend, to challenge, both they recalled more and the quality of the writing was, was definitely better. [01:23:23] Speaker B: Human agency, they were not the tool [01:23:25] Speaker A: agency at the core of what we. And we have agency. We have agency. We're not living in a simulation. This is not pretend. You're not your own main character energy. Like you have agency to actually shape the future as you want it to be now. Some things you can't change, yeah, fine, But a whole lot of things you can. And we haven't exhausted the list of things we can choose yet. Once we have, then we can start worrying about stuff we can't change. [01:23:48] Speaker B: James, that's good, that's good, mate. [01:23:50] Speaker A: Pleasure. [01:23:51] Speaker B: Thank you for the time. [01:23:52] Speaker A: Thank you for like having the conversations genuinely, mate, like, I think that's the thing. We're just having the conversations. Stuff comes from that and you don't even know what's going to spark someone to go, I hadn't thought about that. [01:24:03] Speaker B: Now that's the beauty of a dialogue. I've noticed that this trip, actually you're one of five actual in person conversations have come up here specifically to record. And you notice the difference in the dynamic. [01:24:14] Speaker A: Yeah, right. [01:24:14] Speaker B: It's just, it's very. And the creative energy. [01:24:17] Speaker A: Everything I talk about the lost seasons of COVID deliberately because for me this always carries. It's great catching up with you on, you know, on zoom and calls and whatever else, but whenever I'm down where you are or you're up here, it's always just a totally different. Different level. Right? [01:24:30] Speaker B: Yeah. [01:24:30] Speaker A: Awesome. [01:24:31] Speaker B: More of that, please. [01:24:32] Speaker A: Brilliant. [01:24:33] Speaker B: Thanks, mate. [01:24:34] Speaker A: Thanks, mate. Hi there. [01:24:35] Speaker B: If you're enjoying our content, then why not consider becoming a paid supporter of our work? You can do that at either Substack or Patreon, and the link for both are in the show notes for this episode. If you do become a supporter, then you'll get access to exclusive content, early release content, and also you'll be helping to fund all of the offline work that we do as well all of the youth camps and the events that we speak at, and all that other stuff that happens that you don't see online. A huge thank you to all of our paid subscribers. It's thanks to you that this episode is made possible.