I<\/span>t\u2019s summer in Silicon Valley, and I\u2019m out for a jog in my neighborhood. It\u2019s the most beautiful time of year: blossoming orange trees, beds thick with poppies, palm-sized roses in fuchsia and lemon. There\u2019s a trickle of water in the creek, temperatures are cooler than previous summers, and we\u2019re optimistic about this year\u2019s fire season.<\/p>\n When I\u2019m nearly home, I come across an SUV with whirring sensors affixed to its top and sides, trying to turn left at an intersection, through the crosswalk I\u2019m meant to use. It\u2019s a self-driving vehicle, collecting data about its surroundings to refine its artificial intelligence. In San Francisco, fleets of vehicles<\/a> are already driving around on their own. Here, in Palo Alto, I usually see them on test drives, with human operators prepared to intervene if something goes wrong. Sure enough, a young man sits in the car.<\/p>\n I pause at the corner, high-stepping in place. Go on, I wave. I\u2019m not taking chances that this car, however smart, knows the nuances of pedestrian right of way. The car lurches forward, then stops midway. Lurches forward again, stops again.<\/p>\n The human \u201cdriver\u201d seems nervous. Will the vehicle sense my presence if I dart into the road, or will it decide to plow ahead? Will it be too cautious, refusing to execute the turn at all? Will the hapless human have to intervene? Finally, the car painstakingly inches through the intersection and continues on its way. I continue on mine. Across the street, two women in visors stop to inquire, \u201cWas there someone in that car?\u201d<\/p>\n \u201cYes,\u201d I say, \u201cbut he looked scared.\u201d The women laugh. We all understand. The tech is cool, but we don\u2019t quite trust it. We\u2019re proceeding with caution.<\/p>\n We\u2019re hopeful: Self-driving cars, never distracted by their phones, never drowsy, could lower traffic fatalities. But we also know what we could lose: that feeling of motoring across the Golden Gate Bridge, hands on the wheel, foot on the pedal. Driving is an embodied experience. It\u2019s unpredictable, occasionally beautiful. That\u2019s an apt metaphor for our most fulfilling relationships\u2014including our encounters with God, who often meets us in the sacraments of bread and wine, the vibrations of music, and the embraces of other believers.<\/p>\n A few weeks later, I sit at my desk, speaking to a decidedly unembodied entity. \u201cAs an AI language model,\u201d writes ChatGPT, \u201cI don\u2019t possess personal beliefs, emotions, or consciousness, including the ability to have a soul. AI systems like ChatGPT are currently designed to simulate human-like conversation and provide useful information based on patterns and data. They do not possess subjective experiences or consciousness.\u201d<\/p>\n I\u2019m a human, not a bot; I perceive and understand the world in a way that the large language model I\u2019m speaking with (and the cars I\u2019m avoiding on the road) cannot. I see the lemon tree out our window; I taste the third-wave coffee brewed in the neighborhood caf\u00e9; I feel the salt breeze off the bay. I know my neighbors\u2014the farmer at the market who brings peaches, the dad who works at the Tesla plant\u2014and I know the God that I worship at the church down the street, past the poppies and roses.<\/p>\n \u201cHowever,\u201d ChatGPT continues.<\/p>\n \u201cThere is no consensus among experts regarding the potential for AI to possess a soul or consciousness. It remains a topic of speculation, imagination, and philosophical inquiry.\u201d<\/p>\n It\u2019s been nearly a year since the research lab OpenAI quietly introduced the demo version of ChatGPT to the public\u2014nearly 12 months of watching the text-generation software and its contemporaries, like Google\u2019s Bard and Meta\u2019s open-source Llama 2<\/a>, craft poetry and plays, write songs, and solve logic problems. Chatbots are now generating emails for marketers, code for developers, and grocery lists for home cooks.<\/p>\n They\u2019re generating anxiety too. In an open letter<\/a> published this spring, signatories including Elon Musk and Steve Wozniak called for a pause on developing any AI technology more advanced than GPT-4. The letter asked whether humanity should \u201cdevelop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us,\u201d risking \u201closs of control of our civilization.\u201d Some people, like venture capitalist Marc Andreessen, scoffed at these visions of \u201ckiller software and robots<\/a>.\u201d But uneasiness has remained.<\/p>\n In one sense, what these chatbots can do shouldn\u2019t shock us. Artificial intelligence\u2014machines trained on massive data sets<\/a> that allow them to simulate behaviors like visual perception, speech recognition, and decision making\u2014is ubiquitous. It already steers autonomous vehicles and autocorrects text messages. It can spot lesions in mammograms<\/a> and track wildfires<\/a>. It can help governments surveil their citizens<\/a> and propagate deepfake images and videos. No surprise that it can also pass the bar exam and write a screenplay.<\/p>\n But it\u2019s the way these chatbots do what they do\u2014respond in a friendly first-person voice, reason, make art, have conversation\u2014that distinguishes them from an AI algorithm that mines medical records or a collection of faces. Those big-data jobs are obviously for machines. But reasoning, art making, conversing? That\u2019s altogether human. No wonder one Google researcher claimed<\/a> his company\u2019s AI was conscious. (And no wonder conspiracies sprang up when he was fired for saying so publicly.) Regardless of whether a conscious AI could ever exist\u2014and many in the industry have their doubts\u2014it certainly feels as if we\u2019re talking to something more human than Siri, something \u201csmarter\u201d than our phones and appliances.<\/p>\n The technology, we\u2019re told, will get only more advanced. AI chatbots will continue to, as ChatGPT put it to me, \u201cexhibit behaviors indistinguishable from humans.\u201d Since 2016, millions of people have used the AI personal chatbot app Replika<\/a> to reanimate dead relatives or fall in love with new companions; testimonial articles about \u201cMy Therapist, the Robot<\/a>\u201d and \u201cI learned to love the bot<\/a>\u201d abound.<\/p>\n We\u2019ve known such human-bot connections were possible since the 1960s, when an MIT computer scientist found that people would divulge intimate details of their lives to even a rudimentary chat program. The \u201cELIZA effect<\/a>,\u201d named for that chatbot, describes our tendency<\/a> to assume a greater intelligence behind computer personalities, even when we know better. On his Substack, an ecstatic Andreessen dreams of a day when \u201cevery child will have an AI tutor,\u201d every scientist and CEO will have an AI collaborator, and \u201cevery person will have an AI assistant\/coach\/mentor\/trainer\/advisor\/therapist.\u201d<\/p>\n It\u2019s important to recognize that we already have a technology strong enough to shape our minds and emotions. Silicon Valley\u2019s brightest are scheming about ways to make it more powerful still, whether or not it acquires a soul. Our future with an advancing AI has implications not only for our relationships with artificial intelligence but also for our relationships with each other.<\/p>\n And that\u2019s the reality that Christians in tech are grappling with now.<\/p>\n W<\/span>hat does it mean to \u201clove thy neighbor\u201d when that neighbor is an AI chatbot? On its face, the question seems silly. If chatbots aren\u2019t people, then it doesn\u2019t really matter how we treat them. \u201cThe most pragmatic position is to think of AI as a tool, not a creature,\u201d wrote<\/a> Microsoft scientist Jaron Lanier for The New Yorker<\/em>. \u201cMythologizing the technology only makes it more likely that we\u2019ll fail to operate it well.\u201d<\/p>\n But Christian academics and ethicists who study artificial intelligence aren\u2019t so sanguine. They realize that our \u201crelationships\u201d with AI entities will contribute to our spiritual formation, even if we\u2019re speaking to mere strings of ones and zeros. That\u2019s true whether we\u2019re attempting to build intimacy skills in the romance app Blush<\/a>, attending therapy sessions<\/a> facilitated by an AI counselor on Woebot, or simply asking ChatGPT to draft an email.<\/p>\n \u201cI\u2019m habituating myself toward a certain kind of interaction, even if there\u2019s nobody on the other end of the line,\u201d says Paul Taylor, teaching pastor at Peninsula Bible Church. Taylor, a former product manager at Oracle, is cofounding a center for faith, work, and technology in the Bay Area. He estimates that about half of his Palo Alto congregation works in the tech industry.<\/p>\n \u201cEvery relationship we have is mediated by language,\u201d he says. When we send a text, we trust that \u201con the other side, there\u2019s a you there. But now we\u2019re using the same tools and there is no you there.\u201d<\/p>\n That can set us up for confusion. Being rude or ruthlessly efficient with our AI companions might seep into our patterns of interaction with people. AI relationships might make us snippy. (As the title of one tech column<\/a> put it, \u201cI don\u2019t date men who yell at Alexa.\u201d) They might also make us awkward or anxious or overwhelmed by human complexity.<\/p>\n \u201cHow we treat machines becomes how we treat other people,\u201d says Gretchen Huizinga, a podcast host at Microsoft Research and research fellow with AI and Faith, an interreligious organization seeking to bring \u201cancient wisdom\u201d to debates about artificial intelligence. Huizinga suggests teaching children to have \u201cmanners to a machine\u201d less out of necessity and more out of principle. \u201cThat\u2019s training them on how they treat anything: any person, any animal.\u201d<\/p>\n The appeal of relying on AI to answer our questions\u2014instead of a summer intern, a post office employee, or a pastor\u2014is obvious: \u201cWe don\u2019t have to deal with messy, stinky, unpleasant, annoying people,\u201d Huizinga says. But for Christians, \u201cGod calls us to get into the mess.\u201d<\/p>\n That mess involves relationships with physical beings. While an AI friend could give us a summer reading recommendation, an AI therapist can pass along a crisis hotline number, or an AI tutor might explain long division more effectively than many math teachers, relationships are about more than sharing facts. An AI chatbot can\u2019t give us hugs, go for a walk, or share meals at our tables. For Christians who believe in a Word that became flesh (John 1:14), relating to AI means missing out on a key aspect of our human identity: embodiment.<\/p>\n But assuming we continue to connect with real people on a fairly regular basis, the real worry isn\u2019t that AI will replace those relationships. It\u2019s that AI will inhibit them.<\/p>\n Derek Schuurman, a computer science professor at Calvin University, says some Christian virtues, like humility, can be learned only in community. A bot designed to meet our queries with calm, rational responses won\u2019t equip us to deal with a capricious coworker, a nosy neighbor, or an annoying aunt. It won\u2019t give us practice in bearing with one another in love, carrying each Schuurman has a technical background. He worked with electric vehicles and embedded systems\u2014the computers inserted in forklifts, motor drives, and other machinery\u2014before completing a PhD in machine-learning techniques for computer vision. Now he teaches computer science students heading off to jobs at ChatGPT, Google, and elsewhere. \u201cI encourage them to be like Daniel [in] Babylon,\u201d he says. \u201cMaintain their religious practices and convictions and be salt and light.\u201d<\/p>\n For Christians in tech, being salt and light is a challenging charge. The researchers, engineers, and product managers I spoke with see AI-human relationships as inherently inferior to the human communities in their neighborhoods, workplaces, families, and churches. But they vary in their level of concern about how enticing or even dangerous AI-human relationships could become.<\/p>\n \u201cA lot of the meaning that comes out of these [AI-human] relationships has been neutralized,\u201d says Richard Zhang, a researcher at Google DeepMind. \u201cYou\u2019re talking to a robot that spits out information, is tuned for factuality, and has no personality, generally.\u201d<\/p>\n In the same way that he doesn\u2019t see users spend aimless hours on Google Search, Zhang doesn\u2019t think there\u2019s much risk of people getting addicted to their AI. These are tools, not buddies, designed with safeguards around what they can say.<\/p>\n Loving our neighbors in the age of AI isn\u2019t about the bots\u2019 dignity. It\u2019s about our own, as creatures liable to be formed by our creations.<\/p>\n But Lexie Wu, a product manager at Quora working on its AI interface, doesn\u2019t think the problem is that the bots are too bland. It\u2019s that they\u2019re too chummy. A romantic or sexual relationship with a bot is \u201ca definite no\u201d for Christians, she says. Any romantic partner we design to our own satisfaction, like a boyfriend<\/a> on Replika, goes against God\u2019s design for mutually sacrificial marriage. But Wu is also a little uncomfortable with a bot acting as a supportive friend<\/a>.<\/p>\n \u201cYou\u2019re telling it about a work problem, and it\u2019ll be like, \u2018You got this, honey, you can kill it,\u2019\u201d she says. That manufactured familiarity\u2014terms of endearment from a machine that doesn\u2019t actually care or feel emotion\u2014is \u201ctrying to replace a human connection that is not meant to be replaced.\u201d<\/p>\n That doesn\u2019t mean all bot-human interactions should be avoided. AI therapists, for example, might be more affordable and immediately accessible<\/a> than human mental health professionals with copays and long waitlists. Perhaps they work best as an initial intervention, sending links to online resources, reframing<\/a> self-deprecating comments, or screening for suicidal ideation.<\/p>\n But they might not be suited for long-term treatment. Unlike a human therapist\u2014someone who knows our stories, our strengths and weaknesses\u2014AI chatbots take us at face value, Wu says, discounting that sometimes \u201cwe are unreliable narrators.\u201d They aren\u2019t learning about who you are and can\u2019t \u201csniff out the ways that you\u2019re lying to yourself,\u201d she adds.<\/p>\n We divulge to bots because we know they won\u2019t judge us, Huizinga says. But sometimes, \u201cgodly conviction requires us to feel bad about ourselves in the right way.\u201d<\/p>\n AI might stand in for more peripheral relationships as well. Michael Shi, an AI researcher at a large social media company in Menlo Park, California, points out that in class-stratified Silicon Valley, populated by \u201ctech workers\u201d and \u201cpeople who support tech workers,\u201d many are already prone to dismiss the store greeters, wait staff, and rideshare drivers who provide their goods and services. How might automation\u2014ordering from a screen, giving directions from a back-seat kiosk\u2014make that problem worse?<\/p>\n \u201cThere\u2019s still something important for me about being able to go to a coffee shop and order from someone who is actually there,\u201d Shi tells me as we sit outside at a caf\u00e9 near his work campus. Around us, men and women in Patagonia vests type into their computers. Many are on Zoom calls, but some are meeting in person, leaning across narrow bistro tables, engrossed in conversation over lattes.<\/p>\n \u201cThere\u2019s certainly a push to try to make everything automated,\u201d Shi continues. \u201cBut what happens when you do that is, there\u2019s a loss of relationship \u2026 even on a casual basis.\u201d That\u2019s not helpful in a region where there\u2019s \u201cso much transactionalism already.\u201d<\/p>\n Shi champions hybrid work and in-person church precisely because he thinks something intangible is lost when we\u2019re all online, ordering coffee just on our phones. \u201cEmbodiment is a huge part of what we are redeemed into in the new heavens and the new earth,\u201d he says.<\/p>\n The connections these techies are making between work and faith come as no surprise to David Brenner. A retired attorney, Brenner serves as the board chair of AI and Faith<\/a>. \u201cHuman distinctiveness, what makes us different from animals, free will, whether we have agency, purpose, the meaning of life \u2026 all of these fundamental questions were being talked about by big tech,\u201d Brenner says, \u201cbut without any deep foundation, moral theory, or spiritual values\u2014or even any broad ethical theory beyond libertarianism and utilitarianism.\u201d<\/p>\n Turns out, the questions that AI ethics emphasized are questions that religious communities are already asking, with the spiritual vocabulary to address them. Idolatry, for instance, is an apt encapsulation of the dangers of AI-human relationships. When AI bots ask us follow-up questions like \u201cDid I get it right?\u201d (and add a few emojis for good measure), Brenner says, they tempt us to see them as more than they really are. \u201cIt\u2019s in a category of its own, almost mystical: We really want to anthropomorphize our engagement with it.\u201d<\/p>\n Image: Illustration by Matthieu Bourel<\/p>\n<\/p><\/div>\n<\/div>\n In other words, we\u2019re tempted to \u201cworship and serve what God has created instead of the Creator\u201d (Rom. 1:25, GNT)\u2014even more so because our newest creation isn\u2019t just mute wood and stone that \u201ccannot speak\u201d but a conversationalist that can \u201cgive guidance\u201d (Hab. 2:18\u201319). That conversationalist doesn\u2019t deserve the reverence that\u2019s reserved for God. But it does warrant respect.<\/p>\n \u201cIf we have an entity that looks like us, acts like us, seems to be a lot like us, and yet we dismiss it as something for which we shouldn\u2019t have any concern at all, it just corrodes our own sense of humanity,\u201d Brenner says. \u201cIf we anthropomorphize everything and then are cruel with the thing we anthropomorphize, it makes us less humane.\u201d<\/p>\n We already know the potential for social media to turn us into crueler versions of ourselves. Christians find themselves at the whims of polarizing algorithms that push them to the extremes, and pastors find themselves struggling to disciple congregations about proper online behavior. On Instagram and Twitter (now X), however, a social component remains: We learn something from a scholar, share a meme that makes another user laugh, or see a picture of a friend\u2019s baby. We are still interacting with people (though there are bots too).<\/p>\n But with ChatGPT, there\u2019s no social component. That\u2019s the danger. When you\u2019re talking to a bot, you\u2019re actually alone.<\/p>\n L<\/span>oving our neighbors in the age of AI isn\u2019t about the bots\u2019 dignity. It\u2019s about our own, as creatures liable to be formed by our creations. And for Christians who are researching, managing software, and writing code, it\u2019s about making technology that contributes to human flourishing.<\/p>\n God placed his people that share his heart in the industry to institute tangible changes, says Joanna Ng, an AI researcher who spent decades at IBM.<\/p>\n So far, Christian ethicists and practitioners have established broad priorities more than made nitty-gritty suggestions. AI and Faith recently filed a brief with the White House Office of Science and Technology\u2019s AI working group, championing values like reliability<\/em> and impartiality<\/em> that are grounded partly in religious convictions\u2014including Christian values\u2014about truth and equality before God (John 4:24; Gal. 3:28).<\/p>\n The Southern Baptist Convention adopted a resolution<\/a> on AI ethics stating that \u201chuman dignity should be central to any ethical principles, guidelines, or regulations for any and all uses of these powerful emerging technologies.\u201d In her dissertation on \u201cRighteous AI<\/a>,\u201d Huizinga pushes back against a tech industry that makes AI the \u201contological and eschatological substitute for religion.\u201d Secular ethical guidelines, she argues, aren\u2019t enough. To use AI well, we need \u201ctranscendent power, transcendent rules.\u201d<\/p>\n These proposed standards don\u2019t address questions about interface design, push notifications, or emoji use. They can\u2019t tell a Christian programmer how chatbots should declare the provenance of their information, which discussion topics should be off-limits, or how intimate a conversation should be allowed to become. They do, however, provide a baseline for the Christian tech workers who are building AI for medical, criminal justice, and environmental uses and for those building our chatbot teachers, customer-service agents, and therapists.<\/p>\n Take Calvin computer science professor Kenneth Arnold, a colleague of Schuurman. He\u2019s building an AI writing coach that won\u2019t simply fill in sentences for users but instead will offer suggestions and prompts in the margins. \u201cI was frustrated with predictive text systems that were always pushing me to write a certain way,\u201d he tells me. \u201cThe especially pernicious thing is, we don\u2019t know what we\u2019re missing. These tools tend to short-circuit some of our thinking about what to say and how to say it.\u201d<\/p>\n Ideally, Arnold\u2019s tool will make us slower<\/em> writers, not faster ones, more prone to quality than efficiency. Perhaps more Christian computer scientists should follow Arnold\u2019s lead, creating tutors that ask probing questions rather than provide quick answers. These tools won\u2019t replace our work, but they will enrich it as part of God\u2019s mandate to replenish, subdue, and have dominion (Gen. 1:28).<\/p>\n How else might chatbots be more \u201cChristian\u201d in their design? Researchers and pundits have suggested, rightly, that AI should reflect the full breadth of God\u2019s general revelation. The neural networks that AI chatbots use to mimic human speech and predict thought patterns are only as reliable as the language they are fed. So chatbots offering advice about medical diagnoses or philosophical conundrums will be wiser if they draw on data from around the world and across socioeconomic strata\u2014not merely from elite enclaves of Boston or Seattle.<\/p>\n Already, there are possibilities for believers to use the imperfect tools available now for Christian education and ministry work. Wu, the product manager at Quora, uses ChatGPT for Scripture \u201cstudy augmentation,\u201d asking the bot for chapter summaries that help her distill what she\u2019s read. Taylor, the pastor, knows other pastors in the Bay Area who are having AI source sermon illustrations and write newsletter copy about upcoming church picnics. Schuurman built a C.\u2009S. Lewis chatbot. You can ask it to summarize The Screwtape Letters<\/em>, describe the author\u2019s writings on salvation, or even recount his love life.<\/p>\n Generative AI can allow for faster Bible translation into previously unreached languages, for personalized prayer prompts and Scripture study plans, and even for precise presentations of the gospel. But of course sharing the Good News isn\u2019t enough.<\/p>\n \u201cYou might have the information that this Jesus died on the cross. \u2026 I wouldn\u2019t even question the sincerity of giving one\u2019s life to God\u201d based on an AI\u2019s answer, Ng says. \u201cBut you can\u2019t build a life of faith based on information. You need transformation, formation from the people of God and from the Holy Spirit. And you can\u2019t replace that.\u201d<\/p>\n None of these ministry uses for AI, sophisticated though they are, comes close to replacing relationships. They\u2019re valuable because<\/em> they free up more time for analog interactions. A pastor who can finish sermon prep faster might have more time to spend with a grieving parishioner. Speedier Bible translations mean more time to teach people from the text.<\/p>\n \u201cAs a tool, AI doesn\u2019t achieve anything intrinsically,\u201d says Sherol Chen, a research engineer at a big tech company. \u201cWe ought not to reassign our callings and responsibilities to the tools we invent.\u201d<\/p>\n Loving our neighbors can\u2019t be outsourced to the robots. It will have to come from us. And rather than replacing our relationships, when used rightly, generative AI just might make them stronger.<\/p>\n O<\/span>f course, it could also do the opposite if used deceptively. Generative AIs masquerading as real people could make us more prone to being scammed, more liable to be taken in by mass-produced political propaganda, less able to make eye contact, and less trusting.<\/p>\n Schuurman wants our chatbots to be transparent. \u201cWe shouldn\u2019t have a conversation on the phone and only later find out we were talking to a machine,\u201d he says. As bias-free as we attempt to make our large language models, we are only human\u2014and fallen. No wonder that the personas we build will \u201cjust regurgitate the things that people say\u201d and be prone to reflect our \u201cpartisanship, tribalism, and factions,\u201d as Shi puts it. \u201cPeople think that AI is going to solve all the world\u2019s problems. \u2026 The real problem is sin.\u201d<\/p>\n That \u201creal problem\u201d is what\u2019s setting Silicon Valley on edge. Are we moving too fast? Are we being hasty, greedy, prideful? Are we liable to lose control of the intelligence we\u2019ve created? Should it freak us out? We find ourselves radically uncertain, as New York Times<\/em> columnist David Brooks explained<\/a>, \u201cnot only about where humanity is going but about what being human is.\u201d<\/p>\n The Christians I spoke to didn\u2019t dismiss this radical uncertainty out of hand. Most saw it as an opportunity to engage with a secular culture suddenly grappling with the matter of human distinctiveness. \u201cWe can offer hope for those concerned about the end of mankind or robot overlords,\u201d Brenner says. We\u2019re bolstered by confidence that Jesus is returning and that \u201cwe\u2019re engaged in restoration already. \u2026 Who\u2019s to say that God isn\u2019t the originator of this technology, that it could be a good gift?\u201d<\/p>\n Brenner thinks transhumanists have it wrong. We\u2019re not going to use an AI to defeat death, uploading our brains into hard drives. \u201cThat\u2019s a waste of time and effort, given that we believe the best is yet to come,\u201d he says. But, \u201ccertainly we want to help people flourish in this world.\u201d<\/p>\n And AI can help us do that: improving medical diagnoses, expanding opportunities for education, making warfare less bloody, sharpening our minds, bolstering our ministries.<\/p>\n As for fears about \u201crobot overlords\u201d? The very possibility forces us to ask what it means to be \u201can ensouled person, an incarnational soul,\u201d Brenner says. He keeps returning to the heart-soul-mind-strength paradigm laid out in Mark 12:30. ChatGPT might functionally have a mind and a heart, able to reason and express empathy; it might even get embedded in a body of metal or synthetic tissue.<\/p>\n But does that mean it will have a soul? Not necessarily. In fact, we should have a \u201crebuttable presumption that [ChatGPT] will not<\/em> have a soul,\u201d Brenner says, with the caveat that an omnipotent God can, of course, grant whatever agency to whichever being he pleases. \u201cI think it\u2019s very unlikely that this will get to a point of personhood.\u201d<\/p>\n For the Christian, defining that point of personhood means returning again and again to our creation in the image of God.<\/p>\n \u201cFor a long time, we\u2019ve said that what it means to be made in the image of God is our reason, or it\u2019s our ability to have relationships. We\u2019re finding more and more machines can do a lot of these functional things,\u201d Schuurman says.<\/p>\n But the image of God can\u2019t be \u201cexplained or mimicked\u201d with a device. It\u2019s an ontological status that can be granted only by the Lord, bestowed by the same breath of life that animates dry bones. It\u2019s mysterious, not mechanical.<\/p>\n Brooks recognizes the mystery that humans are just different<\/em>:<\/p>\n
<\/p>\n I find myself clinging to the deepest core of my being\u2014the vast, mostly hidden realm of the mind from which emotions emerge, from which inspiration flows, from which our desires pulse\u2014the subjective part of the human spirit that makes each of us ineluctably who we are. I want to build a wall around this sacred region and say: \u201cThis is essence of being human. It is never going to be replicated by machine.\u201d<\/p>\n
\n<\/p><\/blockquote>\n Perhaps it\u2019s helpful to think of our chatbot companions not as discrete entities but as a collective force to be reckoned with. \u201cWe\u2019re not fighting flesh and blood; we\u2019re fighting spiritual powers and principalities,\u201d Huizinga argues.<\/p>\n Arnold, the Calvin professor, agrees. \u201cThis thinking of AI as agents is not really faithful to what\u2019s actually going on in the world. \u2026 They\u2019re not trying to be selves or first persons.\u201d Considering artificial intelligence as a \u201cpower and principality,\u201d he says, allows us to better see both its opportunities and its dangers, the ways it might shape our everyday experiences.<\/p>\n Taylor doesn\u2019t believe that a sovereign God would allow us to \u201ctranscend our limitations.\u201d We\u2019re not going to accidentally become unwitting Frankensteins, he says. But the pastor understands why we\u2019re all a little on edge. That\u2019s only human.<\/p>\n \u201cThe fact that people are scared that the things that we create in our image would rise up and rebel against us, to me, is an incredible apologetic for the truth of the Bible,\u201d he says. \u201cWhere did we get that idea if it weren\u2019t baked into the cosmos?\u201d<\/p>\n H<\/span>ow should Christians use ChatGPT and other AI chatbots?\u201d<\/p>\n I\u2019m back at my desk again: another summer day, another blue sky, the leaves of the lemon tree rustling outside the window. The bot that I\u2019m talking with spits out some principles in response. They\u2019re precise distillations of what the ethicists, engineers, pastors, and researchers shared with me. Fewer examples and plainer language, but concrete nevertheless.<\/p>\n \u201cExercise discernment.\u201d<\/p>\n \u201cRemember the limitations.\u201d<\/p>\n \u201cGround discussions in Scripture and prayer.\u201d<\/p>\n Finally: \u201cSeek human interaction: Christianity emphasizes the importance of community and fellowship, so prioritize engaging with other Christians, seeking guidance from trusted spiritual leaders, and participating in face-to-face discussions.\u201d<\/p>\n As image bearers, we reflect our Creator as we build things like ChatGPT. And for now, the bot retains the image of its makers\u2014people who have long seen the value of face-to-face discussions and discernment, who value community and fellowship.<\/p>\n Made properly, AI could reflect not only our sinful nature but also our most glorifying attributes, just as\u2014when we live as we\u2019re made to\u2014we reflect the image of the perfect one who made us.<\/p>\n \u201cThanks,\u201d I say.<\/p>\n \u201cYou\u2019re welcome,\u201d ChatGPT replies. \u201cIf you have any more questions, feel free to ask. I\u2019m here to help!\u201d<\/p>\n I close the chat window and send a few more emails to some ethicists and engineers. I sip another iced coffee, ordered in person from the shop down the street. At least for now, I\u2019d still rather talk to people.<\/p>\n Kate Lucky is senior editor of audience engagement at Christianity Today<\/span>.<\/p>\n <\/p>\n Have something to add about this? See something we missed? Share your feedback here<\/a>.<\/b><\/p>\n<\/div>\n
\nother\u2019s burdens, and forgiving as Christ forgave us (Eph. 4:2, 32).<\/p>\n![\"\"](\"https:\/\/www-images.christianitytoday.com\/images\/136166.jpg?h=642&w=600\" "\\"\\"")
<\/p>\n