| I’m Christina Sterbenz, and this is Bloomberg Opinion Today, a Minority Report fan’s interpretation of Bloomberg Opinion’s opinions. On Sundays, we look at the major themes of the week past and how they will define the week ahead. Sign up for the daily newsletter here. As a journalist swallowing a heavy dose of reality on a daily basis, I gravitate toward science fiction (and fantasy) whenever I read or watch for pleasure. So when my worlds start to collide, I pay attention. And this week had a lot of that. When US Transportation Secretary and acting NASA head Sean Duffy (because Trump officials do more than one job now) recently issued a directive to build a nuclear reactor for the moon by 2030, Liam Denning had the same question I did: “Why?”
“We kind of need to build them here first,” Denning says. “Just getting a nuclear reactor that is cost-effective, reliable and ready to go in the United States within the next five to ten years is a moonshot in itself.” Then again, a couple of the nuclear power stations we do have below the stratosphere just got clogged with jellyfish because of warmer ocean temperatures, according to David Fickling. The dark irony of climate migration causing damage to clean energy facilities feels very Vonnegut.
To be fair, Duffy noted lunar nuclear power isn’t a new idea, even for the government. He just wants to move past the research phase. If you want to know how that could go, just read Isaac Asimov’s The Gods Themselves. Speaking of the moon, the space race has long been one of many points of contention between the US and China — and great fodder for sci-fi plotlines. (The Lunar Discovery series has been on my too-long to-read list for a while.) But the bigger contest is now AI, and Hal Brands notes several ways the US could lose. “If the US leads in high-end innovation, China excels at broad adoption — at integrating AI into the everyday workings of the economy and the state,” he writes. Reading that, what springs to mind is China’s robot cops, like the ones in iRobot, Chappie and well, RoboCop — I could go on. They can walk and talk like humans and apparently do front flips. But American police are using AI, too. Officers now have two tools, DraftOne and Field Notes, that can help write reports. Seems potentially problematic because, as F.D. Flam notes, “One of the more subtle and insidious threats posed by artificial intelligence and related technology is its ability to tamper with memories.” Anyone remember that Black Mirror episode? What’s next, pulling up ChatGPT during interrogations? OpenAI may be working to curtail the sycophancy of its chatbot, but the latest version hasn’t solved the problem entirely, according to Parmy Olson. A few people have even gone full Ex-Machina because of it: “A Silicon Valley venture capitalist who backed OpenAI … appeared to have spiraled into delusional thinking after starting a conversation with ChatGPT about an innocuous topic like the nature of truth, before going down a dark rabbit hole,” she writes. All of this is happening at a time when the US is embracing AI in bigger ways — even ones that were previously off-limits. For example, the government had prevented Nvidia from selling its H20 chips to China, largely over national security concerns — until now. “You might wonder what’s changed,” says Dave Lee. “Nvidia has agreed to pay the US government a 15% cut of its revenue of the H20 chip in China. That’s a pretty unprecedented deal.” The Trump administration also put out a new action plan for AI, which outlines the role public policy can — and arguably, should — play in this accelerating revolution. The Bloomberg Editorial Board says it’s great but has one key oversight: The text doesn’t mention the word “immigration” — strange, considering how many of today’s AI leaders were born abroad. “Yet the US makes it needlessly difficult to retain such talented workers, subjecting them to forbidding rules, extended wait times and pronounced uncertainty. (Rampant deportations don’t help),” the editors write. In fact, there’s evidence the US is using AI to crack down on immigration, creating more uncertainty and fear. ICE has tapped into an AI-enabled camera network, according to 404Media. Agents are using a facial recognition app, also according to 404Media. And one CBP agent was caught on video wearing Meta smart glasses, according to — you guessed it — 404Media. Of course, there’s money to be made with all this adoption. JPMorgan is looking into OpenAI, and other investment banks will soon follow suit with private coverage: “With the US stock market at a record high, investment banks are keen to arrange secondary share sales for the hottest unicorns,” writes Shuli Ren. TL;DR: We’re trying to put nuclear power on the moon when confused jellyfish are ruining the ones on Earth. Law enforcement is using tools straight out of Minority Report to arrest people. And not only are unicorns real, but Wall Street wants to invest in them. We really are living in a sci-fi novel. I think I need some fantasy content now ... Bonus Science Non-Fiction: What’s the World Got in Store? Note: Please send book and movie recommendations and feedback to Christina Sterbenz at csterbenz@bloomberg.net. | 
