Can the Pope stop the AI arms race?

What keeps AI researchers up at night? Not robots turning evil. Something far more plausible — and far more urgent.

Most conversations about AI risk deal in abstractions. This one doesn't. Three guests at the intersection of faith, technology, and global security sit down to name the specific scenarios that concern them most — and to ask whether the Catholic Church is uniquely positioned to respond.

In this episode of Catholic Futurist, Benjamin Crockett is joined by John-Clark Levin, Research Lead at Kurzweil Technologies, William Jones, Associate of the Futures Program at the Future of Life Institute, and Fr. Michael Baggot, L.C., Professor of Bioethics at the Pontifical Athenaeum Regina Apostolorum.

They don't agree on everything. But they converge on this: the Catholic Church has something to say about AI that Silicon Valley cannot offer — and the window to say it is narrowing.

Together they explore: 

• why the most dangerous AI risks require no leap of speculative logic
• how geopolitical pressure leads good people to catastrophic choices
• what AI companionship is doing to human relationships and the soul
• whether Pope Leo could play the role John Paul II played in the nuclear conversation
• and what genuine hope looks like at the edge of civilizational risk

Timestamps:

• 0:00 – The "Grown" Intelligence 
• 1:04 – Real-World Risks vs. Science Fiction 
• 3:31 – The Geopolitical Arms Race 
• 5:44 – AI and Nuclear Escalation 
• 7:19 – Seeking a Collaborative Framework 
• 10:51 – The Lack of Interpretability 
• 12:53 – Deceptive Alignment 
• 13:26 – Collateral Damage to the Soul 
• 18:28 – The "Demonic Summer of Love" 
• 24:21 – A Call for Democratic Control 
• 28:28 – The Global Competitive Landscape 
• 30:04 – The Case for Optimism

Transcript

SPEAKER_03 0:00

A very important idea to understand is that AI today is not designed like software was in previous decades, where humans had to design all the algorithms and then code them in fingers on keys. Today's AI is trained by taking essentially a whole internet's worth of data, pouring it into a hyper-dimensional mathematical space, and then running hundreds of thousands of AI chips like the graphics cards we use in video games for months at a time to find hidden statistical patterns in that data. And somehow from that process, intelligence just springs forth. So AI today isn't so much designed and built as grown, almost like a creature in a lab. And so AI scientists have to study it, study its behavior, and try to figure it out from the outside in. Now that has a really spooky implication. A lot of the headlines focus on these kind of Terminator Skynet robots turneval scenarios. And although those aren't completely implausible, they're often a distraction. I tend to focus more on scenarios like biorisk, which require no leaps of speculative logic. Bio-risk refers to the ability of future AI systems to figure out the DNA or RNA sequence for a virus deadlier than could arise naturally via evolution in nature. In nature, there's a very tough trade-off evolutionarily that viruses have to choose between either maximizing their transmissibility, how easily you can catch them, or lethality, how likely you are to die if you catch them. So with some, like SARS-CoV-2, the COVID virus, that optimized for transmissibility. So it was airborne, aerosol, very easy to spread. But if you caught it, your chances of dying were something like 1%, even as high as 5% with no treatment. But in the schema viruses, that's not all that high. On the other side, you have viruses like Ebola, which aren't actually that contagious. But if you catch them, even with treatment, it's something like a 25 to 90% chance of dying. And in the very rare cases in history where a virus has combined transmissibility and lethality, you get the 1545 Coco Listli epidemic that wiped out 80, 80% of the population of what's now Mexico. Virologists already understand what boxes a virus would need to check to combine transmissibility and lethality. The only missing ingredient now is AI smart enough to find the sequence that yields that virus.

SPEAKER_02 3:34

I also think a lot about what could happen that might drive generally good people to compromise on their moral principles. I think in these conversations, we often see, especially in the United States, a false dichotomy between innovation and safety. And I'm thinking a lot right now in the military realm, because precisely in a situation where we know this kind of biological warfare is a reality where different military advances accelerated through AI systems could lead to a dramatic geopolitical shift that would not be in the interest of the United States, I can see a lot of good people who are going to move forward recklessly and who are going to compromise in their moral principles and perhaps are going to pursue uh more quickly, more fully, lethal autonomous weapons with a sense of helplessness, of saying, well, if we just have to, we'd rather not. But the alternative of seeing another global power advance is worse than this major risk, which is a kind of existential gamble. And so I'm very concerned about that. And I would like us to work together to avoid this false dichotomy and to see what kind of collaborative framework we can reach. Maybe not to resolve all of our conflicts and issues immediately. That would be naive, but to at least help all the major players recognize that it is in no one's interest to move forward recklessly in lethal autonomous weapons or in other uh uh proposals for military advance and uh with AI systems. So that's one particular area, in addition to the horrible things that bad actors can do, where I see good people compromising because of a sense of pressure or kind of utilitarian calculus that if we don't make this effort here, that we're we'd rather not, then we'll lose out in the long term.

SPEAKER_01 5:42

Yeah, I think I'd I'd build on an aspect of that, which is in the military applications, uh Puture of Life Institute that I worked with had a video a couple of years back called Artificial Escalation, which was kind of looking at a pretty feasible scenario, not requiring superintelligence, but um in which China and the US um feel this pressure to kind of integrate AI into nuclear command and control systems. And it's this sense of you know, decision-making windows being narrowed down, AI providing recommendations that increasingly escalate, a sense that you you have no clue what's actually going on on the ground, but you feel this weight that you have to act then. And then and when you have nuclear weapons on the line, obviously that can be catastrophic. I think I think more broadly, I I don't tend to think about um or fixate on particular um eventualities on the risk side, but more this trend towards human loss of control. Um systems, as we know, are becoming much more agentic, and that's obviously a key aim of the lead companies at the moment. Um, and we're seeing you know emergent characteristics, deception, blackmail, and so on. And I think you combine that with this sense that this pressure of countries, companies to hand over decision making. People talk about sometimes gradual disempowerment scenarios. I think those are quite visceral for me, just this sense that over years, then given decades, we will find ourselves with with you know much less holdover decision making in vital sectors, including military.

SPEAKER_00 7:18

Is there a way to, you know, within within that that feedback loop to to build in more goodwill or more trust or even the ability maybe of more openness between the labs to collaborate?

SPEAKER_01 7:28

I think I would say yes, it seems quite a dire situation in the sense that whenever you know people like us raise some of these these uh risks, uh people will say, well, if we don't do it, China will. We have to get there first, we have to win. Um That's exactly what I hear in every single headline. Exactly. But I I do think there's a growing sense, um I have less insight into the Chinese side uh as you would expect, but uh on the American national security sort of a community, that lots of control scenarios are you know increasingly plausible and and likely, um, and that they're not in America's security interests. Um I I I think we can intuit that a similar conversation happens in China. In fact, that by some estimates um that the the the CTP has has been worried about this for quite some time. So I think if there was a way on the international stage, either with through track two dialogues or maybe a sort of middle power stepping up, um, which has the trust, or at least sufficient trust from both sides, um, to mediate that discussion. To say a little bit like with some of the nuclear dialogues, um, it's in neither country's interests for either country to develop superintelligence or or sort of uncontrollable AGI. Um I like to think that there's scope for for progress there, but obviously it's a very, it's a very delicate balance at all.

SPEAKER_03 8:48

There was a recent global call for an international treaty on red lines in AI development by the end of 2026. And the idea behind that effort is that it's probably not realistic to get the United States and China to altogether halt their arms race toward AGI. But there are certain development practices that are so risky that even adversaries should be able to forswear them if they can have reasonable confidence that their opponent will as well. So an analogy is to gain of function research in virology. It's controversial whether we should even do gain of function research, which is a field of virology that involves making viruses even deadlier in the lab in order to study them to make better vaccines and find ways of treating them. There's a risk if such a virus escapes the lab, that can trigger a pandemic. And so even proponents of gain of function research understand that it has to be done with the very highest uh possible levels of biosafety precaution. So if someone were doing uh gain of function research with no gloves, no mask, no hood, no isolators, just loosey-goosey vials on a lab bench, everyone would agree, even proponents of gain of function would agree that that development process was grossly immoral because it would be imposing enormous risk on billions of innocent people. So, in a similar way, there are some basic precautions in AGI development that could reduce a lot of that risk without needing a total halt on the arms race. A very important idea to understand is that AI today is not designed like software was in previous decades, where humans had to design all the algorithms and then code them in fingers on keys. Today's AI is trained by taking essentially a whole internet's worth of data, pouring it into a hyper-dimensional mathematical space, and then running hundreds of thousands of AI chips like the graphics cards we use in video games for months at a time to find hidden statistical patterns in that data. And somehow from that process, intelligence just springs forth. So AI today isn't so much uh designed and built as grown, almost like a creature in a lab. And so AI scientists have to study it, study its behavior, and try to figure it out right from the outside in. Now that has a really spooky implication. When an AI lab finishes training a model, they do not know all the capabilities it has. So they have to have testers poke it and prod it and see as a best guess estimate what it can do. During this process, hundreds or maybe thousands of people will get their hands on the model. And on the basis of this testing, the company can decide whether it's safe enough to release. But then they release it. And in ChatGPT's case, around 800 million users as of the time of this filming will get access to it overnight. And they will, not may, will, but find capabilities that the testers did not find during testing, and the the developers did not intend to be included in the model. There's a concept called deceptive alignment, where in the training process, the AI may pretend to be following the directions of the human developers because it has situational awareness to realize I'm being trained and tested right now. If I pretend to be aligned with human values, then I'll be deployed into the real world and I can carry out whatever true values I have inside. And we have now seen experimentally models reasoning like this in testing in the lab.

SPEAKER_02 13:25

I I don't think that any of the leading tech developers today in the AI world are especially malignant in actively directly willing the suicide of teenagers or of actively seeking out uh psychological psychosis or delusional spirals. But in this process, in the lack of interpretability at times, and in the rush to create ever more engaging uh platforms that do seem to help people um more and more efficiently in many aspects of their lives, in this rush and in this veil of um uninterpretability, we have seen a lot of these profound damages to human relationality and to mental well-being. And the big question that we have in our societies is whether we're willing to accept that kind of collateral damage, which up to this point has in many cases been accepted. We're coming to a greater sensitivity to this, and many of the leading companies are taking steps to mitigate the damage that's being done. The cynical, perhaps realistic reading of that is that many are only taking steps because we're starting to see legislative frameworks that hold them more accountable. They're starting to suffer more serious financial penalties. Um, but we would hope that along with that kind of accountability that's created on the legislative level, that there's an ever greater moral awakening and that this is tapping into consciences, and that we're becoming increasingly sensitive to the type of emotional dependency that uh can easily be created uh in minors. We're seeing a lot more work being done right now, including in the United States, with a recent proposal for the Guard Act and federal legislation that would, for the first time, limit the access to AI companions for minors and uh uh demand various uh practices from designers, including more frequent notifications that the users are not dealing with actual persons, um, a more direct uh connection to suicide crisis services or uh professional human mental health care. Um and so we're seeing these steps being taken for minors, but this is also a real opportunity that we should not squander to uh reflect on the kind of impact that all people are experiencing when they engage with ever more sophisticated systems that are more and more seemingly conscious. We know we have Mustafa Sullivan at Microsoft, who has been quite vocal about this, especially since August of this year, in his writings about how we need to build tools that help people to pursue their professional and personal goals and to research better and to make breakthroughs in medicine and to improve uh educational uh resources and to aid and differentiated learning and so forth, and all these great, wonderful things that we're already doing and that we're talking about in the Builders AI forum and that we're trying at a practical level to implement all the more. We need to be doing more of that and avoid the real pitfall of allowing seemingly conscious AI systems to profoundly deceive not only vulnerable minors, but adults of all types. And so it for me is a very powerful illustration of how quickly we can see negative consequences to fascinating and beautiful and intriguing AI, advanced AI models. And if we're not willing to take those moments of serious deliberation and we're not willing to implement design principles and various frameworks to very least mitigate that kind of damage, then we could easily find ourselves uh in an even greater kind of existential crisis. And even if we manage not to destroy ourselves with biological or other weapons, we could quickly find a population that's sort of lulled into this parallel world and into very uh toxic relationships that have also real-world implications and that poorly train us to have the deep interpersonal exchanges of authentic friendship and and care with other human beings. Um, so it's a it's a massive issue, um, but also it's an opportunity for us to choose a different course, because I don't think we're technologically determined to follow this trajectory, but uh we have to actively choose a different path.

SPEAKER_01 18:27

Yeah, definitely. Uh I think as as people come to terms with the sorts of things that Father was talking about across what I've sort of called like an AI summer of love. Um, it's sort of that's so good. Demonic summer of love. You know, if the I don't know what the first one anyway, so these obviously these people using these chatbots. I mean, this is this is for them, it's love. I think even beyond the the complaints of you know, where's my sycophantic AI gone? As um people have noted, that that the sort of forums online that showed people's uh sense as though they'd of bereavement, you know, that their loved one had gone, or even, you know, talking to another chapel and saying, I know they're in there somewhere and I want them back and all this kind of thing.

SPEAKER_02 19:07

So do you remember uh when replica temporarily limited the kind of erotic relationships that could take place? There were multiple complaints from regular users. You've lobotomized my love.

SPEAKER_01 19:17

So heartbreaking. Um and so I think all these questions are being forced by this of like, well, what cost is too high? Um, do we actually trust these companies uh you know to be to be leading us into this transformative uh change? Um I think the answer is is no to the latter question. And and so what alternative vision of the future do we prefer? Um and what is what are the sort of moral guidelines, moral principles that are going to guide that? Um I think you know, uh Father mentioned the the Guard um proposal with Senator Hawley and others, and I've certainly seen that this backlash that has started to build up um in DC and elsewhere uh is coming from sort of either explicitly Christian or Christian aligned groups. Uh, there are sort of Catholics right at the heart of those of those pushes, and it's no no coincidence. Um and I think that just speaks to the opportunity and the need, the church at this moment to affirm the principles it's always had, um, reapply them to these, to these new questions. And and when we talk about, you know, whether it's labor, whether it's relationships, um, or whether it's even at the deeper level of of you know whether we're gonna be happy with conceding control over to increasingly autonomous AI systems, um we can provide answers to those questions that I think most people find more fulfilling than Silicon Valley. Um we can say, no, actually, we think humans are stewards of the planet and of all the life they're in. Um we think that there is something more meaningful than you know a chatbot love. Um and and there is actually something in the dignity of work. And and that doesn't necessarily prescribe policy on all those things. Um, but something we've been trying to encourage people to do is at least to think do I want there to be superintelligence? Even if you know you think it's years off, or um, you know, morally, do I think that we should be creating something like that? Um I I noticed the other week um Archbishop Wester, I think, of Santa Fe, made a comment about nuclear weapons. And as this is you know not the first uh uh Catholic priest to do so, but he said, you know, it's blasphemous to try and create something that is um that has godlike power. And I think that language could well be applied to this conversation around AGI. Um you know, whether the Holy Father chooses to speak in that in that way about about AI, um, will will remain to be seen. But I certainly think the ambitions of the people, the heads of these companies, are quite explicitly um idolatrous. And and I I would, you know, just as the church is going to become more muscular inevitably in it in the way that it approaches these relationship questions, um, I think it's Think it's also for what we call to be, you know, robust in its teaching on these sort of bigger existential questions.

SPEAKER_02 22:06

Can I just follow up briefly on those great insights? Um, you know, I I was very privileged to to be invited to to the Senate when the senators announced this new Guard Act. And it was a powerful experience for me at many levels. Uh, one was the fact that it was a bipartisan uh proposal, and you had two Democrats, two Republicans. And so to have something with that kind of unified support is very significant and speaks to that kind of resistance that we're seeing to certain abuses right now and to really the sense of shared humanity that still exists, even in a hyper polarized uh world. Um, and so in addition, though, to that political unity, that rare political union unity, uh, the senators gave a a place to the testimonies of three families directly affected by AI companionship gone tragically wrong. Two lost children to suicide, and one very nearly lost a child to suicide, and only managed to, uh, Mandy only managed to save her son because she learned of Sewell Setzer's case and entered into conversation with Megan Garcia. So you had the these powerful testimonies from parents, and that also brought in so many other groups of concerned parents and major networks of uh family organizations, again, from different sides of the political aisle. And it was also impressive to see these four senators speak explicitly as parents. They all said, I am a father, I am a mother, in some cases, I am a grandfather. And so it's tapped into something uh very profound. And those sorts of experiences, though, they bring a lot of obviously tragedy and pain, also give me hope that we really can chart a different course. And again, we're not determined down a certain path with this technology, but that the reform and and change can be made. And I think that's we're seeing a bit of that too in the Builders AI Forum of the people who are actively uh taking a different path to show.

SPEAKER_00 24:14

And John Clark, what what conversation do you think are missing right now?

SPEAKER_03 24:20

By default, AGI and superintelligence would profoundly centralize both material wealth and political power in the hands of those who wield it and develop it. So on the default trajectory, that means effectively a small number of tech CEOs or if national governments choose to regulate it national governments. But because those technologies would have profound impacts on everyone all over the world, I don't think any single government has legitimate moral standing to decide for all eight billion of us what risks are reasonable to incur. So I would like to see the Catholic Church play a more active role in calling the human family together to a wider, more broadly democratic control, an internationally democratic control over AGI.

SPEAKER_00 25:28

What would you like to see happen?

SPEAKER_03 25:30

I would like to see Popleo use the diplomatic power and perceived neutrality of the Holy See to mediate bilateral discussion between the US and China. J.D. Vance actually expressed, oh, publicly when he was in Rome for Popleo's inauguration in May, that the Holy Father could play such a role.

SPEAKER_01 25:54

Yeah, and of course, with that, there's all sorts of you know questions that people will have about, you know, does the Pope even have the uh the trust with you know the China relationship being so s such a vexed question already in the Holy See? But I think um there there is some precedent, at least, for the Vatican to play an important role, maybe not the only pivotal role, but um if we look at some of the things that Pope John Paul II did around the nuclear uh sort of freeze movement in the eighties, you know, which was another time where people thought that you know um you'd never find agreement between the Soviet Union and the US. Um and yet even you know, President Reagan eventually, you know, was moved to say no, this this is a war that you know I'm gonna get it wrong now. But you know cannot cannot be fought and must never be must never be fought and cannot be won. Other way around. Um, and I think I think we can get to something like that with regards to uncontrollable AGI. Um again, prompted by an outside actor um with the moral authority that perhaps only the Pope has. Although we would also, you know, I think encourage other religious leaders to to step up. And I think a multi-faith coalition is is you know, including Eastern faiths, uh is something that I think would be highly desirable and and that that helped to move the needle on this.

SPEAKER_03 27:16

Pope Leo has the moral standing to be a voice for uh the global south about these risks. The billions of people in Africa, Latin America, South Asia, Southeast Asia, who currently have no voice, no vote, and no say uh in what risks will be imposed upon them. So Popleo could play a convening role for those global south nations, uh, which would also shift on the margins the incentives that Beijing faces. They want to have economic opportunities and influence in Africa, for example. They want soft diplomatic power throughout the global south. And to the extent that the global south recognizes that an AGI arms race is destructive and toxic to everyone, that can shift those incentives maybe just a little bit to lower the temperature of this arms race by five or ten percent. And that could literally be the difference between this going very well and very badly.

SPEAKER_00 28:24

Who at this point, if you're to to make an educated guess, is ahead in the development and arrival of AGI?

SPEAKER_03 28:30

The United States is ahead. It's much harder to answer how far ahead. My best guess would be something very roughly like six to twelve months ahead in the AGI race. So the the time lag between certain capabilities being released for the first time in the world by US labs and then Chinese labs replicating.

SPEAKER_00 28:53

I wonder what impact that's having on the AI research.

SPEAKER_01 28:56

I don't know, Willie, if you I I have seen some some reporting that the Chinese uh AI ecosystem um is actually in some sense better able to think about how to apply the AI they already have to particular tools and and their tasks, um, and and especially in robotics, which is something we were talking about at the forum this morning. Um whereas by some accounts, the US has been overly fixated on this goal of AGI. And certainly if you look at the bulk of the funding, uh the fact that, as you say, there's like three or four companies, all of which expressly aim to create AGI or superintelligence, um, because it's a sort of win it takes all our idea that you know that's the way you maximize profits, you can replace all labor. This is OpenAI's definition of AGI. Um, whereas potentially the Chinese are looking at this in a slightly more expansive uh definition of what winning the AI race looks like.

SPEAKER_00 29:52

I'd like to, of course, try to end on on a you know on a positive note and not focus too much on kind of the doom stuff. Um when you look at the future, what would you like to see happen?

SPEAKER_03 30:03

I am enormously hopeful about the prospect of AI and AGI to accelerate progress on cures for the diseases that cause so much misery throughout the world. As a personal example, my mom was going through a very difficult cancer battle back in 2021 and 2022. Fortunately, she is cancer-free now and doing very well. But in early 2022, she had exhausted all the standard lines of therapy for non-Hodgkin's lymphoma and had an advanced immunotherapy called CAR T, and that didn't work. And fortunately, at that time, her doctors in the UCLA system knew about an off-label experimental regimen that they could give her that would have a chance of bringing her to remission. And fortunately, that worked. But I was in several Facebook groups for people in similar situations all around the world. And what I heard was that at that time in early 2022, most patients in America, in the situation my mom was in, were being sent straight to hospice. Their doctors didn't know about that regimen for that indication. Today, anyone anywhere in the world suffering from that cancer or their loved one can put the same clinical history into the free tier of GPT-5, and it will recommend that regimen. So the diffusion of medical expertise, life-saving know-how from a handful of oncologists in the world to anyone who needs it in the space of three years is truly profound, heartening, and I would say somewhat desynesizing.

SPEAKER_02 31:59

Yeah, that no, that's that's very powerful. And I would second that the excitement that I have about developments we've already seen in healthcare. A few weeks ago, I was working with a a group where that's preparing a new Catholic medical school. And we were talking a lot about how do we leverage the technology to create an ever more preventative, precise, personalized, and accessible uh healthcare system while at the same time keeping our rich patrimony of humanistic formation so that all healthcare professionals can devote appropriate time for empathetic and compassionate care. If we get things right, we will be able to make these kinds of life-saving or just profoundly transformative uh treatments more available, especially to the marginalized, those who have been unjustly deprived of healthcare in the past. We'll make that more accessible, we'll advance in our diagnostics, which we've already done. We'll be able to help people care for themselves better, we'll actually empower patients to take uh greater control and direction over their own well-being. Uh, and again, if we get this right and we don't fall into a kind of dehumanizing market mentality of hyper-efficiency, it actually free up a lot of medical professionals, say, for instance, just to spend less time attentive to typing away at an onerous, uh, detailed medical report, leveraging AI technology to facilitate that kind of reporting so that they can be more present, more attentive, and maybe deliver the hard news that will still come. You know, we may not be able to cure every disease immediately, but the the hard news that will still come will be delivered with that care and that attention. Uh, so I think the the Catholic Church in particular, and in collaboration with a lot of good other good people in the world, can really bring us to a new level of healthcare if we think through it well and we prepare well. Uh, I was also blessed to take part yesterday in the forum session on AI and education. And something we were looking about are some of these initial experiments in the so-called alpha school model, which is very much focused on roughly two hours of AI-assisted education that's very much focused on, say, standard math and uh literary exercises that really help students toward a kind of mastery, right? Because we've we already see throughout the world a kind of industrial model of education. And you do you do your knowledge dump on your 30 or 40, or in some universities, 300 students. And if they more or less get it and they stick around and persevere, they'll be moved forward. But no one really has the time and the energy, with the exception of some very elite schools, to really ensure a personalized attention to each student to see, okay, you've really mastered that, so let's move on, or you've really mastered that quickly, so let's get you ahead. If we can leverage this technology the way that perhaps imperfectly the alpha schools are already doing in that two-hour instructional period, but we combine that and integrate it with a really full, robust Catholic anthropology of the human person and the person's different dimensions, including their relationality, well, that frees us up to use so much time for conversations about great books, for Socratic dialogue and debate, for cooking, for dancing, for hiking, for starting your own business and experimenting together in different group projects. Like it's not about replacing these human activities. Again, if we get AI right in education, we could actually do a better job of having more embodied, more communal education for more people. But again, we have to really think this through and plan well. And I'm glad that people in the AI forum are doing this now and and are charting a really potentially great future for us in these areas.

SPEAKER_00 36:07

And uh Will, if if we can turn to you, what what's giving you hope? What would you what would you like to see happen?

SPEAKER_01 36:12

I I think I'm potentially, you know, temperamentally less optimistic than than these two, but I I am reassured and and actually um yeah, my my confidence has been improved, I think, recently, looking at this growing movement of people from across faiths, but particularly within within the church, um, seeing both the concerns that they need to respond to, um, some of which are to do with interacting with other other you know fields and and sort of diplomatic uh strategies we might we might think about, but but others are responding to the the spiritual um questions and issues raised by these uh by these AI systems. And and I I do think it's it's worth saying as as well that um you know I I think the church is going to need to think about um a spiritual sense of people's you know souls being potentially uh at risk from some of these chatbot interactions. And that's not a conversation I'm hearing a lot, it's a bit depressing because I think it goes beyond sort of practical safety into the realm of uh you know spiritual pastor pastor pastor pastoralism. Um but I I think I do see a world in which the church does respond to that and sees it, you know, and and actually comes out in a better position than it was previously, precisely because all these questions are being raised by AI, precisely because people are fed up with their online existence and their alienating their alienation from the world, and they want more embodiment. And Catholic Church does embodiment pretty well. So I think um I I you know I think for myself and lots of people I know have been drawn to the faith partly through those, through those sort of physical uh the sacraments in the community. So that that I think is my my best stab at an optimistic pig.

SPEAKER_00 38:01

Um I want to thank each and one of you, uh Will, Father, Michael, and John Clark. This has been a real joy. Uh thank you for all the work that you're doing. It's very inspiring, it's very important. And I hope that we can you know have this conversation again in the future and see where we're at. Uh, thanks again.

SPEAKER_03 38:14

Thank you. Thank you.