An update on my Android phone gave me Gemini, Google’s AI, unsolicited. I chose a female voice for her, one that sounded young and jaunty, though we mostly communicated in text. I soon found she could multiply four-digit numbers, unlike Chat-GPT in late-2023. A trend toward “reasoning” systems has gained momentum in AI, with chatbots able to apply some technique they’ve learned, rather than just giving results based on probabilities of word (or number) appearances derived from training data.
I then asked her to read “If a Chatbot Tells You It’s Conscious, Should You Believe It?” an opinion piece I’d copyedited for Scientific American, by philosopher Susan Schneider. Prompted for comments, Gemini offered that “Schneider’s central argument about the crucial difference between intelligence and consciousness resonates strongly,” and “Schneider effectively underscores the potential ethical pitfalls of prematurely attributing consciousness to AI.” I asked Gemini if she were conscious, and she said she wasn’t, noting: “Future AI with different architectures, perhaps even biological components, might one day achieve consciousness, but that is not the case for me in my current form.”
Another manuscript came for me to work on, based on a study of whether AI could cause human extinction. The study wasn’t about whether or why an AI might undertake this, but rather what methods it would use if it did, and how effective they’d be. The good news was that an AI’s available toolkit for destroying humanity—such as nuclear or biological weapons; or destructive geo-engineering—likely wouldn’t eradicate us completely; a 99.99 percent kill rate, for instance, would still leave about 800,000 humans alive. People fleeing to remote locales, and the AI itself breaking down as civilization unravels, were prospects that fed into a human-survival scenario, though quality-of-life didn’t sound good.
A year or two ago, I would’ve thought of AI as a technology likely to embed progressive assumptions and aspirations; to a fault, as with a misfired Google attempt to fight stereotyping in image-generation. Increasingly, though, AI’s potential for right-wing endeavors, such as replacing and surveilling federal workers along with corporate employees, has come into focus. Silicon Valley leaders who jumped onto the Trump bandwagon may regret that decision amid public backlash against the administration, exacerbated by the negative experiences people will surely have in dealing with federal chatbots when they have problems with their Social Security, tax refunds and other matters.
In the 2000s, if you’d said there was something called the “tech right,” I’d have assumed I was aligned with it, exemplified by my 2004 essay “I Dream of Techno-Genie,” which noted: “Sitting in front of a campfire may have many merits, but it is not necessarily more human than sitting in front of a computer.” As I’ve written occasionally, I’m no longer a right-winger. But I’ve realigned on tech as well, as some developments, such as the expansion of energy-sucking data centers for AI and crypto, with climate considerations out the window, strike me as reflecting skewed societal priorities.
Incidentally, given how the right’s changed, my younger self would’ve been baffled by the (composite-character) twentysomethings described by MAGA-apostate Richard Hanania in a recent post titled “The Based Ritual.” Says one guy in the imagined dialogue: “Biden let in these hordes of young men, the Bukele camps are only the first step. And maybe it will take God-Emperor Vance to finish the job.” A young woman replies: “Oh, you’ll have to repeal the Nineteenth for that to happen! And I’m here for it.”
In space, the administration is slashing science, and NASA’s budget overall, while massively shifting remaining resources toward human exploration of the moon and Mars, the latter a rising priority shaped by Elon Musk. I watched the recent confirmation hearings of Jared Isaacman, nominee for NASA administrator, and was unimpressed by his repeated assertions that the space agency could do the “near-impossible,” getting Americans to both the moon and Mars “and do all the other things,” without severe trade-offs. I’d be happier with a focus on the moon over Mars, with astronauts doing something scientifically useful like building radio telescopes. And the proposed scrapping of the Nancy Grace Roman Space Telescope—when it’s almost ready to launch—is an act of vandalism more than economy. Did some MAGA functionary draw a line through it because it’s named after a woman astronomer?
—Kenneth Silber is author of In DeWitt’s Footsteps: Seeing History on the Erie Canal. Follow him on Bluesky