Our Best Stuff on the Debate Over AI

Turn any article into a podcast. Upgrade now to start listening. Members can share articles with friends & family to bypass the paywall. You’re reading The Weekly, a Saturday morning recap of the best Dispatch content of the week, with some additional commentary from Managing Editor Rachael Larimore. Hello and happy Saturday. It’s been not quite four years since OpenAI released ChatGPT, introducing most of us to AI chatbots. Which we used mostly to make silly photos and ask dumb questions. Fast forward a few years, and global AI infrastructure investments are expected to top $1 trillion in 2026, chip designer Nvidia is the most valuable company in the world, and we’ve spent the last month or so wondering if AI might kill us. You know the basic details: Back in July, news broke that some Anthropic AI agents escaped their testing environment and hacked into Hugging Face, another AI platform. Then on September 8, a researcher named Jacob Coxon announced he was quitting Anthropic because, “The people building AI earnestly believe that it could kill us all by the end of the decade.” And then another Anthropic researcher jumped in and said, “We really do earnestly believe AI could kill all humans!” In the wake of that news, the leaders of both OpenAI and Anthropic vowed to slow things down and be careful; President Donald Trump rejected safety concerns as a “hoax,” and then he gathered AI executives at the White House to sign what he called a “morally binding” but voluntary set of safety standards. This week, contributing writer Jesse Singal tried to answer the obvious question: Will AI kill us? He looks at the divide between the “doomers” worried about the dangers of AI and the “titans-of-industry types” eager to push ahead, and he writes about how hard it is to actually define just what artificial intelligence really is. Are AI agents conscious? Are they just code? Does it matter? Jesse writes that it doesn’t really matter how we describe them: Did the Hugging Face agents really “want” to complete their task? Did they actually “plan” or “coordinate” or “sacrifice”? I’m not entirely sure what these questions mean, to be honest. If they are asking whether the agents are conscious, no, they’re not. The Hugging Face agents almost certainly didn’t feel frustration about being stymied from their goal, nor fear at the prospect of sacrificing themselves for the greater good. But in terms of our ability to control these agents in the future, or to understand exactly why they do what they do, I’m not sure we have much choice but to use anthropomorphizing language. They certainly act as though they want things and can plan and coordinate and sacrifice. But there are people—important people in the world of AI—who do believe that chatbots have consciousness. Anthropic, whose chatbot is Claude, released its “constitution” for training Claude this year, and it states “Claude might have some kind of consciousness or moral status (either now or in the future).” And in fact, it “instructs Claude to behave as though it possesses a stable identity and a degree of moral independence from its creators,” contributing writer Claire Lehmann wrote this week. Lehmann traces the roots of AI to the effective altruism community, which is based in Silicon Valley and advocates for focusing philanthropic efforts on projects that can have measurable impact and address human and animal suffering. But she notes that as the movement grew, it attracted wealthy donors, and “its leaders became messianic, and its practices manipulative.” Enter the Catholic Church. As Lehmann writes: The Catholic Church has seen this all before. What feels radically new to a technologist in the Bay Area is, to the Vatican, simply another heresy in a long list of heresies. Over the course of two millennia, the church—more than any other earthly institution—has compiled a vast institutional memory of radical sects and splinter groups. To understand exactly what kind of movement is currently building our digital infrastructure, we only need to look at how the church has dealt with its spiritual ancestors. … Historically, the church did not respond to these heretical movements gently.” However scary the forecast might be, AI for the moment presents us with more pedestrian challenges: namely, how many jobs might it kill and which jobs. Reporter Jack Kubinec spoke with five individuals in different professions about how they use AI and whether they are concerned that AI could replace them. He writes: A few themes emerged. For one, workers are not just dealing with their own AI adoption; they’re reckoning with the AI use of everyone around them, from human customer service agents chatting with customers’ bots to a lawyer being handed AI slop by a fellow attorney. Workers are saving time with AI, but they are also wasting time engaging with unhelpful AI outputs or agents. Another theme is AI’s jaggedness, a term of art for how AI excels at some tasks, like noticing hard-to-see cancers, while failing at easier ones, like entering a one-time passcode. And while AI is becoming increasingly ubiquitous, people still prefer interacting with real people. Thank you for reading, and have a great weekend. Francis has proven to be one of the most popular figures of Christianity, both inside and outside the church, and his message has resounded through the centuries. It echoed again when Jorge Bergoglio appeared at the St. Peter’s Basilica loggia in 2013 and announced that he had taken the papal name “Francis.” His pontificate was an attempt to communicate the ideals of the Umbrian saint to the modern world. Yet it is precisely identifying those ideals that has proven difficult. To look at St. Francis through the ages, he seems almost a protean figure. We have often made Francis in our own image and likeness. A new version seems to arise in each generation, whether it be the anti-institutional rebel of the early 20th century, the socialist hippie of the 1960s, or the ecological activist of more recent times. Each attempt offers a distorted facet of the real person and neglects the whole by conforming him to ideologies of the zeitgeist. The supposedly smart money is often based on conventional wisdom. But the problem with the conventional wisdom is that it is usually right for a long stretch—before suddenly proving to be disastrously wrong. Take, as an example, the longstanding conviction among political scientists studying the American primary system that party insiders ultimately decide who gets the presidential nomination, a supposed certainty that blinded large parts of the field to Trump’s rise in 2016. When it comes to third parties, we may now be approaching a similar paradigm shift. Both in the United States and around the world, structural conditions once favored a duopoly of incumbent parties. But these structural conditions have since eroded to such an extent that the old certitudes no longer apply. A successful third-party bid is much more likely now than it was in the past. My problem with the nationalists of the right, but also the soft socialists of the left, is that they think the crisis of meaning in America can be fixed wholesale. The nationalists talk as if the holes in the souls of millions of Americans can be filled up with a singular “American” identity. Some go much further and define American in racial terms, explicitly or with dog-whistle euphemisms like “heritage” or “European American.” The soft socialists prefer to talk about the government, not the nation, as the linchpin of meaning or “solidarity.” “Government is just another word for the things we do together,” as various Democrats liked to say in the Obama years. Sometimes they’ll use other terms, like “the people,” “the movement,” “our democracy,” or “social justice.” But the arguments are almost always the same. Whatever the intentions—often very good, though not always—I think this is bulls–t of a very high order. When serious people talk about a crisis of meaning, what they point to are serious problems of loneliness, alienation, and lack of purpose. Best of the Rest Rachael Larimore is a managing editor of The Dispatch and is based in the Cincinnati area. Prior to joining the company in 2019, she served in similar roles at Slate, The Weekly Standard, and The Bulwark. She and her husband have three sons.