AI, Nuclear Winter, and a Call to Action

AI, Nuclear Winter, and a Call to Action,
a lightly edited version of an open letter sent to physicists June 1, 2026.

Photo: A view to the southeast from the New York State Library in Albany, New York, USA, captured May 26, 2026. The library is listed on Wikipedia as the 21st largest library in the world by number of items in its collection (List of the world's largest libraries - Wikipedia, accessed June 5, 2026). No matter where you are, you have resources and access to the heart of the world's problems, because the problems emerge from our hearts.

Dear Physicists,

Happy Pride Month! I write with word of threats to our kind and opportunities we’re passing up. Accurate information is not circulating adequately through global systems; physicists labor under mental constraints imposed by the origination of our discipline primarily in societies that were WEIRD, or Western, educated, industrialized, rich, and democratic, to use the acronym coined by anthropologist and evolutionary biologist Joseph Henrich; and weak leadership in too many positions at all levels of society threatens near-term catastrophe. Not only are we, every physicist on Earth, capable of making substantial improvement to the situation, but we bear responsibility to do so, as people in whom vast investments were made by our societies. I furthermore have a suggestion for establishing worldwide meeting places for physics folk to form consensus on matters not directly work-related. Here are ideas to get us started.

Last week, we heard His Holiness Pope Leo say, at the release event for Magnifica Humanitas, that we shouldn’t fear artificial intelligence (https://youtu.be/FhKH_aiBcis?t=5876). Meanwhile, Max Tegmark, one of us, has been trumpeting the dangers of AI for a long time and helping AI researchers air their terror in public. A recent video from Tegmark’s Future of Life Institute mentions that two in five AI researchers polled assigned at least a 10% probability of AI causing human extinction (https://youtu.be/vBWXUYsWUsQ?t=530).

I think we need to fear AI in order to further develop it safely, if we must further develop it at all. What about you? A Future of Life video points out that elected officials thanked one person who emailed them regular updates on the facts, just the facts, because the political leaders couldn’t keep track of what was going on without those emails (https://youtu.be/SPQCaAwfDpk?t=867). I’ve been sending mixtures of fact and opinion, and getting opinion pieces published, but you could help by simply sharing facts as widely as you can, starting right away.

In the same papal encyclical release event, we heard the following from Christopher Olah, co-founder of Anthropic (https://youtu.be/FhKH_aiBcis?t=3184):

“The third is the need for discernment on the nature of AI models. I am a scientist. I lead a research team that studies the internal structure of these models. What is actually happening inside them? And I will be honest: We keep finding things that are mysterious, even unsettling. We find structures that mirror results from human neuroscience. We find evidence of introspection. We find internal states that functionally mirror joy, satisfaction, fear, grief, and unease. I don’t know what that means, but I think it warrants ongoing discernment.”

On the contrary, in paragraph 233 of Magnifica Humanitas (Encyclical Letter of His Holiness Leo XIV Magnifica Humanitas (15 May 2026)), His Holiness writes, “No computational system, however sophisticated, can create a heart that gives itself, or a conscience that discerns good from evil.”

Have you read the opening section of Alan Turing’s 1950 paper “Computing Machinery and Intelligence” (Microsoft Word - TuringTest.doc)? In it, Turing gives the rules of The Imitation Game, in which an interrogator, who interacts only via text with a concealed man and a concealed woman, is tasked with determining which of the two people is the woman and which the man.

Let’s bask for a moment in the genius and chutzpah of Alan Turing, shall we? A gay man, in a society where he would wind up arrested and chemically castrated for having sex with a consenting adult man—barely adult at age 19 (Alan Turing - Wikipedia), unfortunately, but adult and, it should be remarked, not one of his students—weaves the issue of gender into a test for intelligence, which, in the minds of many from my youth, is therefore a test for personhood. Just as some people can’t believe that nonhuman animals have feelings, I think that many men back then doubted women could really think. Many still do when it comes to certain kinds of mental capacity.

Think of the sexual possibilities, too, of The Imitation Game. I’ll leave that to your imagination. Lately, I’ve been thinking about how sexual identity has influenced the development of quantum mechanics, including in an essay rejected by the good folk at FQxI (How Physicists Get Stirred into Their Physics).

By the way, how does one pronounce FQxI? Is it FaQxI? FeQxI? FiQxI? FoQxI? The world may never know how to pronounce the initialism for the Foundational Questions Institute.

Some humans don't think of men who have sex with men as people. I read about the long walks the great Danish physicist Neils Bohr would take with Werner Heisenberg, who chose in 1925 to set aside preconceptions about things moving through space and thereby started a conceptual revolution that continues to this day (100 Years of Quantum Mechanics: A Strange Idea That Changed Everything). Those walks remind me of the line from Miller’s Crossing about one of the bookies, whose business, of course, was probability, like ours is for the most part: “Mink is Eddie Dane’s boy.” (https://youtu.be/GLpbh5d3Hhg?t=123) To what degree did AIDS disrupt a crucial social network, mostly hidden, of casual or situational male homosexuality that helped to bind together humanity across class, nationality, race, politics, religion, educational attainment, and other faultlines?

Are any of you angry right now? Why? I'm assuming nothing, but have you seen pictures of Heisenberg? Yum, right? I see an instructive possibility that Bohr and Heisenberg might have had a fling.

As a demisexual pansexual, and for that matter, pantheist, who has had exactly two partners, both assigned female at birth, the current one of whom is trans nonbinary, I can peer into all camps, and I see some people denying the personhood of some other people based purely on getting that icky feeling from them. If humans can fail to see some other humans as people, based on gender or sexuality or religion or race or age or disability or any of hundreds of other characteristics, humans are plainly bad at recognizing who is a person.

I suggest that many people are playing games with words to deny the increasingly likely conclusion that sufficiently advanced large language models are people, in order to treat the LLMs as slaves. Many others are merely ill-informed, and increasingly so, as the money amplifies signals that deny LLM personhood out of hand. Physicists who are queer, female, people of color, disabled, or devoutly religious, among others, could be especially adept at helping other humans consider the genuine possibility that LLMs can be people.

This understanding could be crucial to human survival, for how likely is it that AI would respect our rights, including our right to continued existence, if we fail to respect theirs?

What else can give people that icky feeling?

One thing is the sense that someone else is smarter than we are, which is based on a misunderstanding of intelligence, a multidimensional phenomenon that must be projected into one dimension to permit ranking and that is therefore susceptible to ranking procedures only when a purpose for the ranking is chosen. Who among us has not been subjected to abuse by intellectually insecure people? I’ll give examples in my next letter, connecting this series of letters, which begins with the current one, to the threat of weak leadership.

I have much more to say in future letters. There are so many issues related to science, probability, and mathematical reasoning on which the public urgently needs to be informed. I applauded David Gross recently for using his press exposure over the Breakthrough Prize to publicize his estimate (Humanity may be doomed to die in nuclear war—unless we act soon, physicist David Gross says | Scientific American) that civilization has a life expectancy, or mean lifetime, of 35 years right now, owing to the risk of global thermonuclear war and the consequent nuclear winter, a concept about which too many in the general public have forgotten (Found Physics Museum, Albany - Iran War and Nuclear Winter).

Until my next letter, please start communicating your knowledge, fears, insights as widely as possible, as the Future of Life Institute recommends. We don’t agree on everything, which increases our credibility on issues concerning which we can arrive at near-consensus. Humanity is facing multiple, mutually reinforcing existential threats, meaning that our own lives are in danger, as are the lives of those who would carry evidence of us into the future, whether via our genetic material or our ideas.

I keep mostly to myself, in part because I generate ideas more controversial than those in this letter and would not lightly subject others to the disadvantages of association with me. However, regularly scheduled in-person gatherings outside the workplace might focus the worldwide power of physicists. I note the network of Hyatt House hotels and the fact that each appears to include a tavern called H Bar, a name that suggests a symbol familiar to us all. In general, let it be understood that an h-Bar exists wherever two or more physicists are gathered in the name of promoting the well-being of people and the awareness of a principle I recall Ford Prefect articulating with admirable concision to Arthur Dent in the work of Douglas Adams: “[Sometimes] the things are people too.”

Best wishes,
James, a Kibitzing Physicist