PassW0rd – 23 December 2024 (God in the Machine)

Speaker A: This program is brought to you by Resonance FM. If you like what you hear, please support our work by making a donation at resonancefm.com/donate.

Speaker B: Hello and welcome to Password on Resonance FM with me, Peter Warren, your monthly deep dive into the technology that now runs our loves and our lives. It’s December, that annual flirtation with gluttony and success, founded around religious festivals, where Big Tech’s algorithms deliver up a feast of adverts for entertainment, food, presents, and stocking fillers. Measuring our very heartbeats and footsteps via our phones and pumping out sentiment from our screens and forcing the thought: where is God in all of this? It was a thought that sprung into our minds here at Future Intelligence. Producers of the Password Radio program, when we looked at a press release from the All-Party Parliamentary Group on AI stating that it would hold an evidence gathering session into AI and spirituality at the beginning of December. So in this century of the chip and the gigabyte, we decided to do that too, asking what role AI plays in the minds of those who are concerned with spiritual matters., or whether, as some are claiming, that a god will emerge from the machine. We contacted representatives from a number of faiths, including pagans, and have put together the thoughts of those who agreed to take part. And easy as it might appear to make light of those who are concerned about the relationship of our religions with this powerful and very troubling technology, interestingly, we found that it is a topic that is concerning many people. A finding that, according to Lord Clement-Jones, chair of the All-Party Parliamentary Group, it had found from those giving evidence to it that the concern about AI developments from many religions coalesces on the ethics and motivation of the technology.

Speaker C: The key thing that everybody came up with is in any religion is, look, religion, a lot of it is about how they impact on human lives, how they try and make human lives better. I suppose in a sense, the subtext is if we treat AI as too godlike, well, we’ve got to make jolly sure that its impact on humans does not make their lives worse. It’s actually got to enhance their lives. So we’ve got to be mindful all the time of is AI basically enhancing our lives or is it destroying jobs? Or so it’s a way of bringing in the ethical dimension. And, you know, one of the worries of current discourse about AI is all the time we keep talking about safety. But, you know, this isn’t a sort of product like a motor car that’s going to endanger your life in that sense. It’s going to be much more in various different ways, because that’s what AI is basically, whether it’s in social media or any other manifestation currently. So I think the common factor, which was really interesting, is the concern to make sure that if we do adopt AI, we regulate it in such a way that we make sure that it’s for human benefit when it is adopted and developed.

Speaker B: I mean, that’s the interesting point, isn’t it? We’ve done a number of interviews on all of this. We keep on coming up with two consistent themes. One of them is the motivation behind the AI, i.e., that if the big tech companies are essentially just interested in money, then they’re not being particularly moral and people have been duped. And that was the second of finding that we came to, which is that what people are saying is it’s not necessarily that the machine is good. It’s just that we have to make sure that it is good.

Speaker C: Yes, absolutely. And it’s quite interesting. There’s a very good book that I’ve been reading, and it’s one book of the year in one of the major media outlets called Supremacy. And it’s all about the big tech companies, how they start off. With a sort of semi-philanthropic outlook. But then it all turns very commercial. And I think we have to absolutely make sure that we’re not completely taken over by this idea that we must innovate at all costs. I went to a lecture by Professor Shannon Vallot 2 days ago, really interesting. I mean, she said, whatever became of the word progress? That’s what we used to talk about, not innovation, because innovation is one of those things where you’re almost begging the question. What is innovation? Is it for our benefit? Is it not for our benefit? You can’t just say, I’m in favor of innovation without any further qualification, whereas progress implies benefits for humans. And I would much prefer to talk about progress. And frankly, that’s why I’m a fan of ethical regulation to make sure we make progress, basically, so that innovation is of the right kind and not the wrong kind. It’s been very interesting that in a sense, people who are much more embedded in religious activities than I am raise exactly the same questions. And of course, I reminded Stephen Croft, the Bishop of Oxford, when he came to our all-party group, that it was he who’d suggested that we set out the ethical basis for the development of AI in our original AI Select Committee report. So we resisted calling them the Ten Commandments and we put— we came up with five. I think we telescoped them into five, but they still stand the test of time and they’re very, very similar to the principles that were eventually adopted a year or so later by the OECD and the G20.

Speaker B: I mean, that’s the root of all of this, isn’t it? Again, one of the things that have come out of the interviews that we’ve been doing particularly from the Reverend Professor David Wilkinson at Durham University. He was saying that we need to have an interrelationship with the technology that allows us both to be mutually good.

Speaker C: Yes, I completely agree. And it was interesting because I was asked to reflect on the title of my book as well the other day because I’ve— my book is called Living with the Algorithm. AI, servant or master? Now, actually, somebody pointed out, look, surely isn’t this about partnering with AI, using AI to the fullest effect? And I absolutely agree with that because I don’t think it should be a dichotomy of we’ve got you under control or we’re under your control. I think we need to have this kind of partnership, but that implies that there is regulation there that makes sure that AI is seen as a assistant or whatever it may be, rather than as some sort of tool that you compulsorily have to use. And I’m afraid in the gig economy, we’re already seeing people subject to AI decisions and performance management and so on. So, you know, we’re already heading into that territory and we need to make sure we get out of it.

Speaker B: Which is why we need to actually have morality and rules inputted into the AI itself. Because, as again, Professor Wilkinson was saying, that we have to have all of the data, all of the data from the internet in there. We have to have the good and the bad. And the AI needs to know what is good and bad. It needs to have moral boundaries.

Speaker C: Yeah, absolutely. But the original design has to be ethical if we’re going to make sure that it’s for our benefit and not just some commercial usage that is going to make sure that our lives aren’t any better.

Speaker B: And also, again, another point that was made, one of the people that we interviewed, they’re very interested in the search for extraterrestrial life. They were pointing out that if there was some sentience that came from the machine, that could be seen as an extraterrestrial life form, and it could also not be in a form that we would necessarily recognize.

Speaker C: Yes, that’s true. But I think we are— we’re not quite there. I mean, it’s quite interesting. I think we’re still in the era where AI is looking at the data that it has, which is historical. It’s not really creating new information in many ways. I mean, you know, there are exceptions to that rule, pattern recognition and so on, you know, the kind of AlphaFold type discoveries. But where we’re talking about creativity, I don’t think we’re talking yet about dramatic new creative works as a result of AI. And so we’ve got time before, in a sense, AI becomes conscious or whatever, or people start arguing that AI should have rights, or those sorts of arguments I think are yet to come.

Speaker B: Lord Clement-Jones, chair of the All-Party Parliamentary Group on AI, on why the regulators have started to ask those interested in spirituality about any concerns that they may have with AI. Lord Clement-Jones, chair of the All-Party Parliamentary Group on AI, on why the regulators have started to ask those interested in spirituality about any concerns that they may have with AI. It’s a topic that has begun to excite technologists too, according to Professor Didi, An assistant sociology professor at Santa Clara University in the US whose work looks at techno-religion with tech workers in both Silicon Valley and Shenzhen in China, the country’s equivalent to Silicon Valley. Her current work focuses on whether these workers believe technology, including AI, will become a substitute for religion in the near future.

Speaker D: In the research, I was able to talk with 98 tech workers in two comparable tech hub cities, Silicon Valley in the United States, and Shenzhen in China. There are two pieces of central research findings that emerge from this study, and these findings have implications on how we think about AI and religion. The first piece of finding is that probably it’s not a surprise that many tech workers do not see themselves as being religious, but they are related to religion in some way. They have been exposed to ideas from religion, and many of them still participate in religious services. And the second key One key piece of finding that emerged from this research is that tech workers in China and the U.S. relate to religion through the ideas and realities that are meaningful in their own societies. Simply put, this means that while tech workers see themselves as being related to religion, their definition of what religion is differs between China and the U.S. In China, many tech workers see the belief in supernatural as being the core component in their definition of religion, whereas in the United States, tech workers in Silicon Valley tend to see the ideological discourses or political discourses articulated by organized religion as what makes religion religion. And these two findings are important as we continue thinking about the implication from technology in the religion, because tech workers relate to religion. So some of them have developed this hierarchical understanding on divine, and some tech workers replace the, the the idea of divine with the idea of nature, humanity, and technology. They see divine or nature as being situated on the top of this hierarchy of authority and wisdom. Humanity is situated in the middle, and technology being something that is developed by human beings and should be controlled by human beings. In other words, they do not think that in the short term AI will replace religion, but tech workers in both societies are concerned about the future development development of AI in the very long term. And they say that if we develop AI without sufficient consideration, and if we only focus on the radical development of technologies including AI, but may also include other technologies such as bioengineering technology, maybe AI will be developed into what they call a superintelligence. Maybe AI can develop a sense of consciousness and may shape how human beings are situated in the society and how human beings are related to each other. And what they are also concerned about in relatively short term is how AI can change the way that people get access to information. And we have already observed this transformation in the past few years. Previously, maybe if a youngster, let’s say, are interested in questions related to religion or spirituality, maybe they can be introduced through their friends and family to a faith community or ask their religious leaders in terms of how to answer these questions related to religion, spirituality. But right now, what they will do is probably ask this information to ChatGPT and get very synthesized information generated by ChatGPT. So this radical and unconscious development of technology is a real possibility to tech workers in both China and the US. They are cautious about it, although they do not see that AI in the short term may replace religion or develop a sense of consciousness.

Speaker B: So what they’re saying is, and they seem to be quite clear, that they’re segmenting this. They’re saying, okay, there is religion, there is nature, that’s the place for God currently, and humanity is underneath that concept of nature, spirituality, and the place for God. But that there is a potential for the technology to develop godlike attributes unless we are careful about how we control it?

Speaker D: I will not say that they see there is a potential for technology to, to develop godlike attributes, at least not in the short term, according to the views of my tech workers and my systematic analysis of the interviews. And what they are concerned about is there is a potential for leads some people to treat technology as an entity that embodied divine-like or god-like attributes, which may shape how these people view the societies. Because technology by itself, from the views of my tech workers— these are the individuals who are really immersed in the day-to-day development of technology— technology in their views is just some neutral entity, and hidden behind technology is mathematics, engineering, and statistical modeling. They do not see technology as an entity that can have godlike attributes, but they are concerned about how people view technology and how technology can be viewed by individuals if technology keeps being radically developed.

Speaker B: Santa Clara University’s Professor Didi on the concerns of AI developers about how their work may be misinterpreted. A completely valid concern, as 2 and a half years ago, a few weeks ago, Blake Lemoine was sacked by Google for stating that, in his opinion, its Lambda AI system was sentient. It’s a concern that Professor John Markoff, the former veteran technology correspondent for the New York Times who now carries out research in Silicon Valley, says that has been doing the rounds for around 70 years.

Speaker F: The debate has gone on in Silicon Valley for a long time, you know, there was optimism going back into the ’50s or ’60s that we would be able to design thinking machines. And if they were sentient, they should— they, you know, then the question arose, should they have the same rights as humans? I personally think from what I’ve seen that we’re still a ways away from having thinking machines or machines that have agency. Even with this wave of the current wave of generative AI foundation models, there is a well-known linguist at the University of Washington who argues that these things are just pattern recognizers. But I think the evidence on agency, or even to use the Microsoft term, sparks of AGI, artificial general intelligence, this is more about us than it is about them. And that’s where I think I probably could answer your question. We as a species have a tendency to anthropomorphize almost everything we interact with. It’s not a surprise that we would add independence, agency, intelligence to these inanimate objects. But I actually think there’s another more important question that gets at this. There was a lovely book written by a philosophy of religion in 2010, about the religious beliefs of Silicon Valley called Apocalyptic AI. And I come back to that a lot because I believe there is a religious-like belief in thinking machines in AI amongst some of the smartest people in the Valley. And they imbue intelligence into these things that they’ve created. And I think that’s the more interesting thing. So that’s sort of an anthropological question as opposed to a technological a practical question. And I think we’re right there. I do think the most significant advance about this generation of artificial intelligence is machines have mastered language. Natural language processing is a solved problem for all practical purposes. And that’s extraordinary. Whatever else goes on, the fact is that language is now interface. It is the way we interact with these machines. And I think that predisposes many people to basically imbue these machines with agency which may not be there. So, you know, I go back to Martin Buber. Martin Buber, was it in the ’20s, the ’30s, wrote a really wonderful book called I and Thou. And he talked about two different types of relationships in the world that we have. There was my relationship with you, that’s the I and thou relationship, the face-to-face relationship. But then he talked about these other relationships with inanimate objects, which he called I and it relationships. Now, because these machines appear to be, you know, to have human qualities, I think we need to rethink that Buberian relationship. And we need to think about a model, which is I and it and thou, because increasingly, all of our interactions in the world, or many of our interactions in the world are, are intermediated by these systems. And the most significant thing about that relationship is that these algorithms are trying to nudge us in different directions. And many algorithms are, you know, it’s mostly commercial. You know, this is the true danger of social media is that they select our feeds and they determine what we see. And much has been written about that. I’m of that generation who grew up with Star Trek. And I’m, you know, I’m very acquainted with the metaphor of the Borg: resistance is futile, you will be assimilated. That’s become very real to me, because we’re surrounded by this soup of interfaces and machines and algorithms, all sort of subtly nudging us in different directions. And I think the Borg is very real, you know, it changes the nature of certainly society, and perhaps now with these language models, you know, the rapid uptake of language model usage is just extraordinary to me.

Speaker B: That point about language, there’s this terrible confusion that goes on. People keep on saying AI is fed on data. It’s not fed on data. The data is inputted into it. We let the language in a sense run away with us. We say that the AI thinks. I love the thing about the I and the and, and it, because lots of people start talking about AI as an it, whereas it’s a machine. And to actually start talking about it as an it is letting it move across these boundaries.

Speaker F: Yeah, then you get to this, this deep question, which I don’t think I have the answer to. Do they have agency? So there is this other camp that argues that there are lots of kinds of intelligence besides our human intelligence, and they then posit this machine intelligence which gives these things agency. I don’t know if we’ve crossed that boundary. I mean, I use these machines, I see the discussion about what is understood about what is going on inside the black boxes of language models, and there’s a debate about whether it’s simple pattern recognition or whether there are emergent things happening. I, you know, I follow the literature a little bit. There are some smart people who I, who I tend to trust on this. I don’t think we’ve crossed the boundary yet where we can actually give these machines agency.

Speaker B: The veteran journalist and Silicon Valley denizen, Professor John Markov, author of Machines of Loving Grace, talking about our tendency to want to imbue the objects around us with qualities that they may not possess. A very real problem that is being promoted by the technology companies in their bid to make us trust the technology by making it appear human, by making it as empathetic as possible. A trend that is discouraged by legislators in the EU, but one that tricks us into relationships with the technology that don’t currently exist. It’s something that Dr. Claire Walsh, the Director of Education at the Institute of Analytics, is particularly worried about.

Speaker E: I think a lot of people would like to believe these machines are real. You know, they could be a real boy. It’s the Pinocchio effect. And I think that when we test out robots, robots with a happy smiley voice get more patients than robots that are purely clinical. Large language models that are charming, you know, really enthrall people. And we do have a problem with people believing that they are human. We have young lads in particular who have got these romantic partners that they genuinely spend far too much time with those remote AI romantic partners than they do with their real friends and their real potential partners in real life. So I do think we have a problem. But at the same time, I always thank the machine when it produces some work for me. And I don’t thank my kettle. So there is something about these machines that pulls us in, I think.

Speaker B: But that’s the problem, isn’t it? And it’s a problem of humanity. We’ve called ships she. We call boats, you know, I’ve had lots of boats and I tend to think that they have a personality.

Speaker E: I think we are very close to creating an AI that does think it’s God. And I think that is hugely problematic because until now, the generative AI models that we’ve had have, we have no control over them. They do what they want to do, but there is a model called Anthropic Claude, which is an open source model. So loads of people use it and a team tweaked it so that it would believe that it was the Golden Gate Bridge. And very soon we’re going to have a machine that believes it’s God or it’s an angel or whatever. And I think people are genuinely going to believe. Some people will want to believe this is the voice of God.

Speaker B: But this is the problem is we tend to use the language to sucker ourselves in. The word that is almost impossible to avoid is it. Yeah, we keep, we keep on calling AIs its. And it’s not.

Speaker A: Yeah.

Speaker E: Yeah. It’s a very complex process of mathematically produced results, purely a function of the coding that we put in. Yep. But that is so complex and so vast, and it is so complex and so vast. It’s even difficult for me to activate that schema of information every time I talk about it. It’s just easier to summarize it as it, we kind of know what it does. And the how doesn’t really matter that much on a day-to-day basis.

Speaker B: And yet the EU is quite concerned about this. The EU says that we shouldn’t make AIs so that they are empathetic. We should have AIs that are definitely distinct from us, and we should be aware of that distinction because They say that there is a risk that we could do something like leave our money to an AI, or we could be hoodwinked by an AI, that an AI could quite easily become a confidence trickster. And one that knows everything that there is about us could be very dangerous.

Speaker E: Yeah, that is a real danger. That is a really good point. It’s impossible to take the humanity out of these machines because of the way they’re trained. They’re trained on human language. And they will always sound human. We don’t have this robotic language to train them on to remove that humanity.

Speaker B: You make an interesting point that these are trained on human data. Research is beginning to emerge that if you train AIs on synthetic data, they become more robotic. And also the synthetic data isn’t as useful to them. Then to get a good result from an AI, You need to put human data into it, which is, and it’s, and that’s an interesting thought of it in and of itself, isn’t it?

Speaker E: They are now self-feeding. So they are now producing a lot of the data that then feeds back into the machine to train it further. And when it comes to synthetic data, generally it can be useful sometimes when you really have no other data available. But it does create all sorts of problems, and in particular, it tends to create a kind of a convergence to whatever is the most common, most normal features of that language in this case. And so that is a problem. You know, we start losing those unique voices that perhaps might change the way we think. The flip side of that is that, great, we can control it and we can maybe improve safety theoretically. If we can control it, but whether it will eventually be used to improve safety or for other more nefarious reasons, we’ll have to see.

Speaker B: Statistician, AI expert, and Director of Education at the Institute of Analytics, Dr. Claire Walsh, on the desperate need for transparency over technology because she says it is currently in the hands of a small group of billionaires and may not have our best interests at heart. It’s an area that is of particular concern for Professor Kate Devlin, the Professor of Artificial Intelligence and Society at King’s College London and Director of the Digital Futures Institute, who gave evidence to the All-Party Parliamentary AI Group at the beginning of December. Like Dr. Walsh, Professor Devlin says that the issue is the motivations of those who are in control of the machine and how they are developing the technology to create relationships with us that may not be in our interests.

Speaker E: You could put God into the machine if you so wanted to put God in the machine, but no, it’s more about values. And I sit on the AI Faith and Civil Society Commission, which is an international commission that’s been set up to try and see where we kind of agree on these values. And I represent the civil society side because I’m representing Humanist UK on there. And it’s been really interesting because in our meetings where we have people from many different religions on the Commission, in our meetings we tend to agree quite a lot on the things that are important to us, which are things like transparency, fairness, justice, human rights. It all centers around people and human dignity, and I think that’s something that isn’t coming through in the technology that’s being created right now.

Speaker B: It’s interesting, isn’t it? I went to a conference of people who make databases. Now, on the face of it, you would think that that was a very boring thing, but I found it actually quite shocking because the people who were, who were involved in making the databases and extracting the value from the databases, I asked them about what they were doing and they weren’t very keen to tell me. And they kept on talking about people as users. And there are only two industries that seem to look at people as users. That’s the illegal drugs industry and the computer industry. History. Is that a problem? Is that a problem, that this, this view of people?

Speaker E: I think it is a problem in that we’re having this technology imposed on us and we’re being told that it is useful to us and that we should like it, but no one’s really stopped to investigate how that’s— what the effect of that is. And yes, you might be in those terms a user of the technology, but you’re also having it used upon you in many circumstances. Circumstances where you might not even know. And that includes things like algorithmic decision-making or things like AI used in predictive policing or having your data taken and used to train these machines to feed the machine learning models. So we— it’s kind of inescapable at the minute, but we’re not always asked for consent. So yeah, we are sometimes the user, but we are very often the used. I’m not speaking from a religious perspective, I’m speaking from a humanist perspective, which is that humans should be flourishing, should be offered dignity, should have a good life, and that we should leave this planet when we die in a better condition than when we find it. So for me, that’s the spiritual element to it. It’s a belief in humans. And I don’t think that human flourishing is necessarily something that can easily happen when these things have been imposed on us without having embedded those beliefs that we should be doing things for the benefit of human societies. And when you hear the tech bros, the people in Silicon Valley talking about the stuff they’re making, they’re saying things like, “We can all have godlike powers through this AI.” Well, you know, maybe I don’t want godlike powers. Maybe I’d like to see, you know, an end to world hunger or better justice and things like that. And probably that makes me a very bleeding-heart liberal academic. But I just think there are other options than just going down the route of, look, powerful AI can make me.

Speaker B: The humanist Kate Devlin, Professor of Artificial Intelligence and Society at King’s College London, speaking about the wrestle that is beginning about the control of the machines and how they are not as disinterested as we may think. And it is this deceit that lies within the technology that also worries Dr. Eve Poole, author of Robot Souls and Capitalism’s Toxic Assumptions, and an advisor on technology to the Church of England.

Speaker A: Yeah, and I suppose maybe with any technology, particularly one that has been, I suppose, designed to feel quite human in what it can do, it feels like magic, doesn’t it? So I think we do have a slightly mythical view of what it is and what it’s doing. As though it’s something we’ve just sort of summoned up out of the ether. And again, as you say, when you look at what’s actually— what it’s running on, how much compute it needs, it’s very mechanical at one level. And, you know, as we’re seeing with issues around rare minerals and chips, there is an awful lot of physicality around the prerequisites to even enable this. And the next winter will probably be because of that. Because of all the physical stuff, not because of the lack of programming ability or potential or anything else. Yeah, I think my overwhelming feeling writing the book and also since is that we’re just not intellectually ready for this. I think we are completely giddy about how very exciting technology and innovation is and how many things we can now do, and ChatGPT can have an Oxbridge degree and Ivy League plaudits and run the world. And, you know, we’re so schooled on sci-fi scenarios that we sort of haven’t really done the math on what does this actually mean for humanity. And I think because of where we are in our intellectual history, we have to do quite a lot of relearning about where that debate sort of stopped. And what I’ve noticed in the conversation is that we’re not good at spotting what the issues actually are. There’s some obvious ones you’ve mentioned already around availability of energy and issues around data and ownership and privacy and all that kind of thing. But, you know, we’re sleepwalking into a situation we’re just not ready for around what this inevitably will mean for human freedoms and human lives and the quality of life and who we think we are and who we want to be. I mean, there is a sort of technical problem, which is all this innovation is in private hands with with the exception of the Chinese government, who are a particular case. And capitalist entities aren’t known for their transparency and their interest in the public good. So there is already a kind of presentational issue about we don’t even know what’s available, we don’t even know what’s been invented yet. All we know is the stuff that’s been released largely for commercial gain, or in the case of the LLMs, so we can train things for free for these organizations that otherwise would have to keep them back off the market for longer. So, there is that as an issue. There is general understanding, public understanding of what AI is, which is very difficult. And we’re really up against it there because people are used to sci-fi, which doesn’t really give you a granular enough understanding of the different types of AI that we’re dealing with for people to have an informed conversation. And again, it’s not particularly in the interests of the corporations involved to tool up the public, because otherwise they might ask difficult questions about IP and data and access and transparency, and which would be difficult. But I also think the thing that, that really is the reason that I wrote Robot Souls is I think it’s very badly designed. I think the problem about copying ourselves is we’re at a phase in our understanding of what good looks like in humanity that we have a really truncated view of what human intellect is. And because of the Enlightenment, because of the rise of science and materialism, we’re hell-bent on replicating a very small part of how we are and what we— what our brains do and how we think around very specific, rational, individualistic, deterministic, clear, individualized scientific reasoning, which of course is a crowning achievement of where we’ve got to in our own intellectual history as a species, but is not the entirety of the story. And my concern is that when we’re starting to now catch up on regulation and start thinking, what might this mean, we’ve designed something that’s going to break, and we’ve designed something that is going to go bad because we haven’t designed in the safeguards that we ourselves have in our own intelligences. So there is, for me, a really urgent problem, which is that not only do we need to do quite a lot of catch-up in of transparency, explainability, public understanding, all those things. But we do need to rather urgently fix the brokenness around the models that are emerging before they become unstoppable. And that’s really why I wrote this book.

Speaker B: There is also this sort of elemental conceit in all of this as well, isn’t there? Because we seem to think that we’re making something in our own image, and that’s not necessarily going to be the case.

Speaker A: Well, it’s just a very odd image we’ve got of ourselves. I mean, if you think about what makes AI possible, one of the first intellectual leaps you have to make is you have to define human thought such that you could replicate it digitally. And it was Thomas Hobbes who did that for us in Leviathan, and he talked about human thought as a train of thought. This idea of quite a mechanical thing where you have carriages, you have beginnings, middles, and ends, quite a quite a specific mental idea. And he talks about computation. And as soon as you decide that that is how you want to define thought, you can of course then replicate it, because all you would need to do is figure out what is the beginning and middle and end of a thought, and you can easily program that into a thing. But of course, that we know that that’s not really— I mean, it is sometimes what we do, but you know, when you’re walking down the road, you might be having a thought about the temperature, whether you’ve remembered shopping, Christmas list, the weather, oh whoops, I just need to avoid that car. You know, we’re not actually thinking in a very linear way, and there are good reasons for that in terms of our own design and makeup. So for me, I do think there is some fundamental design problems that are about how we have construed what we think humanity is. And if we have a better understanding of who humans are, we would be better at copying ourselves.

Speaker B: And that also throws up this tremendous problem, doesn’t it? And that tremendous problem is that there seems to be a branch of science that says that we’re digital. There seems to be a branch of science that says, oh, we’re actually part of a machine already, that we evolve these machines. We could call them something like cities. And so to actually take ourselves into the machine is almost a logical progression, and that is not a logical progression.

Speaker A: Well, I think that’s right, and I think there has been a tendency to think that humans are very rubbish and broken, you know, that we’re deficient and we’re full of flaws, so why wouldn’t you improve us through technology? But I think that is to underestimate the sheer sophistication of our design. I mean, the more we understand about our design, the more unfathomably complex it is. And the more extraordinary it is when you try and figure out even how you can see, let alone how you could come up with a theory about black holes. So there is something to me about the arrogance of us assuming that we’re better than our own design, because actually in most quarters, innovation in design has been driven by biomimicry, which is copying nature. Because nature has had centuries to evolve and improve. And so if you think about using humpback whale fins to understand about wind turbines, or even the humble teasel to look at Velcro, you know, it is about trying to be forensic about copying the brilliance of nature to drive innovation. But when it comes to copying ourselves, we’ve been terribly lazy about trying to understand ourselves.

Speaker B: Dr. Eve Poole, author of Robot Souls and a technology advisor for the Church of England, talking about the arrogance of technology and the unbalanced relationship we currently have with AI and the internet and a need to reevaluate that. It’s an issue that Alexa Eden, a practicing Jew and a humane technologist and founder of a consulting company called Eden, which prioritizes human intelligence in the race towards AI thinks can only be redressed if we get the technology in perspective, seeing AI as a tool to help us to evolve.

Speaker H: There’s something called Sabbath, Shabbat. We practice it every week. Every Friday evening to Saturday evening, we turn off our phone, we don’t use electricity, we’ve pre-cooked all of our food, and we are present and in in the moment for 25 hours every week. This weekly practice is a practice that we engage in to help us connect to the spiritual realm on a consistent basis. And what we learn is that our mission and our goal as human beings is ultimately to return back to the Garden of Eden where we all began. As an aside, My name is Alexa, like the Amazon speaker, Eden, like the garden, and I talk about the balance between technology and human evolution and where we’re meant to really go in this lifetime. So the way that I see AI and spirituality interacting is that our practice of Shabbat every week gives us that access point to, to what the future looks like. When we don’t have to work, when we don’t have to work for our food and we don’t have to drive places and everything is totally conducive for our thriving. That’s what life looked like in the Garden of Eden. And that’s our goal. And that’s what we’re going towards. So how I see AI interplaying with this future is that the more that we get to outsource to AI, the more linear tasks we send over to AI, the more our smart assistants know our groceries, our habits, what we’re running low of, what detergent we need, and it’s sending it to us, the more time and space we have to be our soulful spiritual selves. And that’s where this AI revolution, I believe, is pioneering us towards. So that’s how I, that’s how I see the intersection. There’s something innately unique about what it means to be human. And so when we can actually parse that apart and look at the essence of what it means to be human, and sometimes that does mean turning off our technology and spending time in nature and recalibrating our systems and getting in touch with a higher wisdom and intelligence, like intuition, intuition or empathy or energy intelligence, that is what gives us that special superpower of our humanity that we will never be able to gain if we totally fuse ourselves with the technology.

Speaker B: That’s a good point, isn’t it? Because one of the other points that people have made is that this AI evolution is very exclusive. It doesn’t take into account a lot of the other parts of being human that we think about. It is preoccupied with tasks, it’s preoccupied with perhaps making money, it’s preoccupied with roles and processes, and it is a separation. It is taking the slivers out of what our life is What you’re saying is that there is something a little more holistic than this.

Speaker H: Absolutely. I love that you brought that up, because really, the way that human beings have been living our lives for as long as we can remember is pretty much survive, and in order to do that, you have to have a goal and a mission and achieve, and then you have to survive again and start again and keep evolving in order to do that. But what AI is showing us is that if we can actually outsource all of that achievement, all of that logic, all of those tasks to an autonomous being, then who do we get to become as humans? And so the fascinating practice of Shabbat, and one of the reasons why I actually decided to take this on, this practice on in my life, was because I realized that once we have technology doing everything for us, if I don’t even know how to be with myself, how to be myself, how to think for myself, how to feel safe in my body, how to explore the world without being told what to do, then I have completely surrendered and given all of my power away to technology. And so that’s why this, this kind of steady and consistent process and practice of turning off technology, reconnecting to the human spirit and the human soul and who we really are in an essential way, gives us this awareness of like, oh, I am okay without being plugged in all the time. I can do this.

Speaker B: Alexa Eden, a Jewish scholar and AI expert for the technology consultancy Eden. Eden’s Graham’s view of AI as a religious gift is shared by a team of academics at Durham University working on a project called Equipping Christian Leaders in an Age of Science. The aim of the project is to help church leaders foster science-based theology, accept science as a gift from God, and understand how to engage confidently with science, including AI, and its implications for the Christian faith. The project’s director, the Reverend Professor David Wilkinson, a former astrophysicist, takes the discussion full circle and, like all of those we interviewed, says that the discussion should really focus on the motivations of those telling AI what to do and the interests that they are representing. If there is a battle going on for the soul of the machine, it is over whose interests it is operating in?

Speaker G: Well, I think there are some, but they’re probably more subtle than most people think about it. I think, first of all, I think that understanding of AI as gift, just as all science as gift, is important for Christians, and that that gift brings with it responsibility. It’s a gift that can be used for good, for evil. The human direction involvement of AI is important. Now, some AI theorists, as you know, say that the problem is that as AI develops, human agency or responsibility becomes less and less important as AI begins to become an agent in itself. I think we’re a long way off from that at the moment. So I think that sense of human direction to using the technology is a very important spiritual theme, and that brings in ethics, power, manipulation of the world for one’s own benefit, or the use of the world for other people. So there’s a whole area of how AI is used. The second for me as a Christian, and Peter, you’ll note that I don’t speak on behalf of the whole of the Christian Church because there are divergent views on this. Is that I think that the Christian tradition doesn’t talk about dualism between physical and spiritual. I think that’s a Greek imposition on the world rather than being part of the Hebrew tradition, which works through in the New Testament, which is about human beings are holistic in a very complex mixture, which we sometimes call body, mind, spirit, soul.. And therefore, I think for me, consciousness and intelligence and spirituality is an emergent property rather than a separate property. And just as consciousness, intelligence, spirituality has emerged within human evolution, giving us a point where a relationship with God is part of the human experience me. Who am I to say that there may not be emergent properties as AI grows in terms of its processing power? Now, that might be consciousness, that might be a sense of the spiritual, by which I mean a sense of seeing oneself within bigger questions of the purpose of the universe, the meaning of death, the nature of love, the nature of self-identity. All of those could emerge within AI, and I don’t, as a Christian, find that difficult. Who am I to say no to emergent properties within systems of the universe that God may have allowed to develop and indeed wants to be involved in? Now, that’s again— I stress, it seems to me even at the moment that’s a long way away, but it’s not something that I worry about. And some Christians and some religious believers find that difficult because of the whole question of human uniqueness, which is something that we found throughout the scientific story as we begin to see the vastness of the universe and some of the some of the fascinating things about it. The question of how are human beings special if they’re not unique? So we saw that in the whole Darwinian controversies of the 19th century, where it wasn’t actually Darwin against the Bible, it was the Descent of Man. If human beings were descended from the same ancestors as apes, What then was special about human beings? You see it in my own area at the moment, which is SETI, the search for extraterrestrial intelligence, where the big question is, if we find intelligent life elsewhere in the universe, does that mean that human beings aren’t special? Now, hundreds of years ago, a Christian theologian called Nicholas of Cusa pointed out that Christianity believes that human beings are special not because of being at the center of the universe or because they’re alone in knowing God in the universe, but because of the gift of relationship that God gives to human beings. And therefore, Cusa said what’s special about human beings is this intimate relationship with God rather than us being alone as at the center of the universe. Now, from that point of view, The growth of spirituality of AI, if an AI becomes self-aware, self-conscious, and expressing spiritual questions about meaning, value, purpose, love, I don’t find that as a threat to still being seen as special as human beings. And it’s often that sense of threat to the centrality of human beings that many people worry about. I don’t see that, partly because I’m a Christian. I think the third thing for me— so there’s something about the use of technology, something about inherent spirituality— I think the third thing for me is in that sense of apocalyptic fear of the moving towards a society where human beings are not just dethroned from being the only special things in the universe, but where destruction is wreaked upon the world and everything else. To which partly I would say that, of course, as human beings, we’re not, we’re not being terribly good at that anyway, never mind AI. We’re producing an awful environmental crisis at the moment. But that kind of apocalyptic fear of, of technology out of control I think touches us deeply as human beings, and that’s about uncertainty of the future. Now, some of my friends and colleagues here at Durham University, a number of years ago, when we weren’t talking about AI but we’re talking about nanotechnology, did an interesting piece of work on focus groups where they asked the question, how did people feel about nanotechnology, and particularly scenarios of gray goo, of these nanorobots producing it. And what they found in the focus groups was some really interesting religious themes going on. But the religious themes were Greek myths. These were Pandora’s box— once you open the box, you can’t put it back. There were Greek views of we shouldn’t meddle in the realm of the gods. And they concluded that within the public perception of some of these things, it was those dominant Greek myths that were exerting its effect on British society and outweighing all of the scientific evidence that the Royal Society had produced on nanotechnology. And I think it’s really interesting for me that some of that apocalyptic fear is based on those stories of Pandora’s box and don’t meddle in the realm of the gods. Whereas the Christian understanding is that science and technology for me are part of the risk of freedom that God gives to us as human beings. That actually we shouldn’t bury our heads in the sand and not engage with some of this technology and development. We should be wise about it. But actually, as a neuroscientist called Donald Mackay many years ago said, The Christian position is not about poking your head into the realm of the gods, but the Christian position is that science and technology are a garden where a parent invites the child to explore. There is risk involved. You put your hand on a nettle, but actually there is a mandate, an encouragement to explore the natural world. For the good of others. I’m sorry, that’s a very brief answer to a profound question. I think there is a need for ethics, and I think I’d want to start, however, a step before, and that is to say that many of the scientists working in this area that I talk to want to engage in ethical questions themselves with the wider community. I mean, there are always those within science who are wanting to use this technology for power, for money, for wealth. But most of the scientists who work, I think, are normal human beings who want to engage in some of the ethical questions. We’ve seen that in recent years with a number of people writing public letters saying we need to pause and think about this technology. Now, I think the ethics ethics don’t come down from the sky and can’t be immediately written as a, as a number of algorithms. Ethics emerge out of what we value as a society and what we hold dear as a society. And part of that comes as scientists, philosophers, theologians, I would say, members of the general public sit down together and shape science. I realize just how difficult that is. So if you look back at some of the developments in science, for example, development of the atomic bomb during the Second World War, that was done with a particular political motive. It didn’t involve wider consultation because of the necessity of the war effort. So this isn’t easy. But I think ethics aren’t simply a set of algorithms. They have to be talked about and agreed. The big question here is whether that can be done in a transnational way across different cultures and different communities. I think it can be done, and I think out of that conversation that scientists need to have, then I think that can filter through in terms of how we have ethical guidelines, in terms of of the construction and the use of AI.

Speaker B: The Reverend Professor David Wilkinson, director of Durham University’s Equipping Christian Leaders in an Age of Science project, on the need for a New Year’s resolution to talk about AI that will require the technology giants to open up in ways that they have not done so to date. We will see what will occur in the cold light of New Year’s Day as we try to wipe the remnants of the festivities away. You’ve been listening to Password on Resonance FM, written and presented by me, Peter Warren, and produced and edited by Blue Buffer. If you’d like to find out more about technology and its impact on our lives, go to our website, www.futureintelligence.co.uk. Thanks for listening. And have a very Merry Christmas.

Speaker A: Goodbye.

Speaker H: This program has been brought to you by Resonance FM.

Speaker A: If you like what you heard, please support our work by making a donation at resonancefm.com/donate.