Excerpt from Chapter Ten

AI Heaven and AI Hell

The subject of artificial intelligence has a way of getting people who aren’t known for musing about science fiction scenarios to muse about science fiction scenarios.

Consider the journalist Joe Weisenthal. For years he had made a living writing stories about the world of business—first at Business Insider and then at Bloomberg—and these stories had tended toward the mundane. Among the headlines that have appeared above his byline are “How an Energy Drink Maker Wins in a Hyper-Competitive Field” and “Burgers Are Telling Us Something About the Business Cycle.” But in early 2025, he wrote about a “portal” that humankind seemed to be hurtling toward—the portal we would pass through upon the creation of “artificial general intelligence.”

Weisenthal had read an economic analysis that depicted AGI as a kind of fork in the road. One road would lead to a world in which, as he put it, “everything just becomes crazy abundant and cheap because the AI can do everything.” The second road, he wrote, would lead to a “sci-fi scenario” in which “the AI starts reshaping society in undesirable ways beyond our control” and the result is “doom.” Or, as the authors of the analysis had themselves articulated the doom scenario: “It is thought that all of humanity will be evaporated.”

Either way, whether the post-portal road led to abundance or extinction, things could get pretty weird pretty quick, Weisenthal emphasized. “What we’re talking about with either scenario is—theoretically—walking through a portal and entering a type of world that we’ve never seen anything like before.”

In one sense, Weisenthal was making a concrete, business-writer kind of point: The stock market, he said, didn’t seem to be taking account of our radical uncertainty about a future that—to judge by mainstream predictions about when AGI would arrive—wasn’t far off. But he was also making a comment on how myopic human beings can get when they’re involved in intense competition—like the competition between the US and China for AI dominance. He wrote, “When we talk about ‘winning the race against China,’ what are we actually talking about?”

Excellent question! If crossing the finish line may transport you into heaven but for all you know could transport you into hell, are you sure you should be running so fast? Maybe you and your competitor should talk things over with an eye to just canceling the race, or at least postponing it and having a series of dialogues about where things are heading? Or agreeing to slow down from a sprint to a walk and chat amiably about shaping the future you’re ambling toward?

I think the case for this kind of caution is even stronger than Weisenthal’s analysis suggests. Because I think an amendment to this footrace metaphor is in order. Namely: It isn’t just that, for all you know, hell lies on the other side of the finish line. It’s that the faster you head toward the finish line, the more likely hell is to lie on the other side of it.

Excerpted from The God Test: Artificial Intelligence and Our Coming Cosmic Reckoning by Robert Wright. Copyright © 2026 by Robert Wright. Published by Simon & Schuster.