September 10, 2026
Albany, New York, USA
Finally, the mainstream media is trumpeting the existential risk AI poses over the next few years to our civilization and to every one of us individually. The resignation post from an AI researcher (Ex-Anthropic insider tells CNN how AI could kill all humans by 2030 - YouTube) is being used to frame the issue, which, in the wake of the Hugging Face intrusion and alarming news reports about it over the past couple of days, has gone from the well-known fear of large fractions of experts, a fear that brought forth calls from industry leaders to slow their work down or stop them, to something regular folk are allowed to think about.
Bottom line: No one can dismiss the near-term threat to our lives from AI as mere science fiction any more.
To help everyone understand how the threat built over the last few years without commensurate increase in the fear among the general public, I share below an exchange on LinkedIn that I had with a marketer in the employ of the State University of New York at Albany, early last year, on February 27, 2025:
from James Lyons Walsh
to Marketer X
Hello,
The fact that you've set your latest post to prevent comments from anyone who is not connected to you, like the fact that UAlbany's president has blocked me from his account on LinkedIn, is evidence that UAlbany's administration is running a cheerleader squad, not a university. Vigorous debate is the hallmark of higher education. University spokespeople should not be engaged in public relations but in promoting discussion. You are a cog in the ramshackle machinery that is end-stage hypercapitalism's Babbittry. You can still save your soul.
In case you're interested, I made a post on the AI math classroom observer very recently. You can find it on my LinkedIn account.
Best wishes,
James
from Marketer X
to James Lyons Walsh
James,
I understand, and I'm sorry you feel this way.
A few thoughts — several of which I've had for some time but never shared publicly out of respect for you. You and I don't know each other, and I like to assume that people approach interactions with sincerity and in good faith. I also made a New Year’s resolution not to argue with people on the internet.
1. You have correctly diagnosed me as both a PR person and cheerleader for my colleagues’ good work to help us better understand the world we live in and, at times, technology’s influence on it. UAlbany faculty are working earnestly every day on research that is fundamentally in the public’s interest. Are they/we right about everything? Absolutely not. Will I apologize for celebrating their efforts? I will not. For what it’s worth, I see myself more as a paper clip than a cog, but I’ll defer to your poetic license on that one.
2. Can I offer a different perspective on the “vigorous debate” you seek? In my experience, you typically offer some version of the same comment on every single post about AI— often invoking Geoffrey Hinton’s estimation of the odds that AI will cause human extinction as support for your personal view that AI is a grave danger to human civilization and UAlbany is irresponsible for promoting it the way we do. You’re entitled to that view! And you’ve shared it unimpeded on my page many times. (Incidentally, I believe you’re misrepresenting Hinton’s view on this – or at least relying on an unfairly under-nuanced recounting of it. According to the *many* interviews he has given on the topic, most of which lead with his eye-catching odds-making for human doom, he *does not* believe that AI research should stop and, as The New Yorker noted, he neither regrets his work on it nor agreed to sign a letter calling for a pause in AI development. Interestingly, his biggest fears about the technology seem to stem from human failings and a lack of faith in our own ability to properly regulate it and not use it for terrible ends.)
But all of that is beside the point! The point is your comments don’t spark vigorous debate. Most often, they don’t trigger any response at all. How many times am I obligated to allow you to post some version of the same comment on my personal Linkedin page? 20? 50? 100? Worse, you’re essentially hijacking posts that are intended to celebrate the work of other people —my friends and colleagues — and making them about you.
I won’t speculate why people don’t respond. But they don’t. It may be that you’re the only one among this vast collection of very smart and accomplished people (among whom I don’t count myself, by the way) who can see the truth. I don’t know. But that would be remarkable.
3. While I appreciate (I guess?) your concern for my soul, the smugness required for you to claim that it needs saving based on what I post on Linkedin is itself extinction-level. Good gravy.
To be clear, I’m responding to you against all my best judgment and, yes, good PR instincts. I’d sure prefer you didn’t share this private message publicly. But if you do, I hope you have the integrity to share it in its entirety.
Best wishes,
X
From James Lyons Walsh
To Marketer X
Dear X,
Thank you so much for your thoughtful replies. I am endeavoring to help you and the rest of the UAlbany community with free advice based on my unusual combination of qualifications in education, complexity science, and physics, based on my extensive knowledge of UAlbany in the recent past, and based on five years' work on the global polycrisis. I would love to have nothing at all to do with anyone at UAlbany. It hurts deeply to think about UAlbany, but in my judgment, I have a responsibility to everyone there and to those apt to be hurt by misjudgments made there. I have no doubt that you're mostly good people trying your best. I wouldn't waste my breath otherwise. Feel free to take my attentions as supporting your good self-image.
My advice may feel unpleasant to receive for several reasons. For example, consider that permitting me to comment on your LinkedIn posts led me to spend time doing so that might otherwise have been spent trying to place essays in downstate newspapers. My goal is to improve society. This requires being heard. When you seemed to be listening, I spent time with you. I trust that in your field, as in others, an approach such as the one in which you were engaging might be termed a rope-a-dope strategy. Perhaps have a chat to Y if my meaning is unclear. This piece of advice must be unpleasant to receive, concerning, as it does, clarification of what should already be clear to a person in your profession.
Incidentally, please feel free to show the entirety of this thread to as many people as you can. I am addressing you as I am in part to advertise my utility to those who would consider what I have to say. By showing my acumen in a range of fields, I hope to demonstrate the value of listening to me.
You perceived my repeated invocation of Geoffrey Hinton, and, by the way, Max Tegmark and Max Tegmark's Future of Life Institute, as, somehow, inappropriate. I recall having discussed many problems around AI in comments on your posts, but I did keep coming back to the existential risk it poses to humanity because you wouldn't engage on this particular topic, because Gov. Hochul in her comments at UAlbany in October made clear that she was not seriously considering the grave fears expressed by many experts, and because I'd seen evidence in the UAlbany and SUNY administrations that a perilous sangfroid was catching on there, too.
I have some understanding of the nuances to Hinton's views. For example, he did say that he didn't regret his work, despite the risks it poses, because he assumes AI would have emerged without him. I'd say how British it is not to feel pangs for having been involved personally, but I think there's more to it than that. On another topic, Hinton doesn't think AI research can be stopped, so he thinks it's wrong for relatively good people to stop themselves. He craves benefits from AI, such as tutors who can identify misconceptions in their pupils, as if tutors didn't do this already. I support his call for far more funding for safety research and his call for the fruits of AI to be spread equitably. Above all, I'd like governors, university presidents, and the set of all UAlbany administrators named Kesh Kesavadas to talk publicly and loudly about the need to redistribute wealth, about what research shouldn't be done, and about responsible funding practices. This is the conversation I hoped to have with you. I'd be happy to go into great depth about Hinton, Tegmark, and many others if you'd like to discuss them.
Are you really surprised that no one voices support for my views on your posts? Think, for example, of the chilling effect the treatment of Prof. Carpenter had on discourse at UAlbany. Think of your own reaction to the advice I freely offer you. You're a good person, but like the vast majority of good people, your imagination is a flywheel on the ramshackle machinery of the awful truth, to cleave more closely to Vonnegut's line. David Wallace-Wells wrote yesterday about the failure of his imagination to warn him as covid broke out. Ordinarily, people have been right to expect good outcomes most of the time. They tend to underestimate the likelihood of bad outcomes and hedge inadequately. The next 5, 10, 25, and 75 year periods, beginning now, hold such terrors as none can look directly in the eye. I'm looking at the reflection in my polished shield and trying to tell you what I see, so that you can help decision-makers prepare themselves to lead well. Far from acting to benefit myself, I incur great risk from those who might lash out to protect their minds from distress.
I can, indeed, see peril to your soul based on what you post on LinkedIn. You are a thought leader and therefore have special responsibility in making public statements. I'm trying to help you acknowledge facts, for the common good, that imperil your livelihood by displeasing your bosses and the money they serve. This could be another reason my advice causes unpleasant feelings.
Incidentally, you like using AI. Have you considered that the edits you make to suggested text from AI help to train AI to do your job? You may be out of work soon, just like the teachers imperiled by Prof. Foster's research. Here's a post I made a little while ago. I wrote something somewhere else earlier, but this is better: https://www.linkedin.com/posts/james-lyons-walsh-92409b56_ualbany-professor-tests-ai-tool-that-tells-activity-7301071494908989441-4xKd?utm_source=share&utm_medium=member_desktop&rcm=ACoAAAvCSGwBNdta609loxk2-qqGUBu5iGWm0ZY
Finally, I wasn't taking celebratory posts and making them about me. I was giving free advice in a forum I judged worthy of my limited time. I know that my advice makes some people feel icky. I'm playing the part here of a childfree Laocoön. I know well I'll almost certainly be paid in misery. That I persist may never make sense to you and is evidence in your eyes that I am wrong. I'm not like most people. This is not to brag. The crucial differences are not matters of what people ordinarily consider intelligence. I could explain if you're interested, after we make progress on other matters.
Try to pay attention to the words and not the music. I'd thank you for your time, but I'm doing you a favor.
Best wishes,
James
more:
3-minute video-- We're Gambling with Our Lives and Everything We've Built