AI Psychosis: High on their own supply

Ok, I would not be surprised if, like me, you are getting a little sick of hearing about AI. I was working on a new post about banjos when I got distracted by a YouTube interview with Cory Doctorow. It is good enough and fresh enough that you should give it a half hour of your time. Then we can all get back to the fun stuff.

Cory can be relied on for clear-headed analysis of AI and other technology topics. And as a bonus, he is really entertaining. When I saw this in my YouTube feed, I thought it might be another rehash of his book “The Reverse Centaurs Guide…to AI” (by the way, READ IT!). But no, Cory has moved on to new insights about the empire of the tech moguls and what the real technology threats are, (Hint, it’s not the AI Frankenstein).

Cory Doctorow interviewed by Dr Myriam François. Sept 23, 2026

View on Youtube

Some of Cory’s insights, paraphrased by me, follow.

Dehumanism by Design

On the tech moguls: They suffer from “AI Psychosis” being “High on their own supply” they believe their own hype. And it is hype. They believe in the coming apocalypse and are preparing. Like any “preppers” they believe they are wizards and saviors: “If I am good at inventing God, or stopping people from inventing God”, I am either a savior of humanity or a good wizard fighting the evil wizards. They are neither; they are simply “banal monopolists”.

Most of these folks, Cory maintains are solipsists. Because their works are the cause of so much suffering, they must dehumanize their victims. Best example. Like Cory and me, Jeff Bezos is an aging male and is prone to having to pee all the time. If Amazon forbids its drivers from stopping to pee at reasonable intervals, Bezos must decide, consciously or not, that they are less that human.

There are numerous examples, which go a long way to explain the thinking and behavior that Naomi Klein and Astra Taylor describe in their new book. This is the best explanation of why they constantly hype the Godlike power of their AI products, and the existential threat they may pose, regardless of evidence to the contrary. So, they are wizards and saviors.

Demystifying AI

Cory believes that there are many threats facing us by this technology dominated world, but few of these threats are posed by AI. The worse threats come from corporate negligence and powerful technologies in the hands of “dumb-dumbs”.

He starts from the “hugging face” incident. Here, he gives the clearest analysis I have heard. I know you have heard it all already, but Cory adds some insights about how the AI was trained on data from the “capture the flag” game played by real humans at security conferences and that this vaunted intelligence amounts to a series of guesses in a loop. Cory asks, if the AI broke the sandbox our reaction should not be “wow, you really made a powerful superintelligence” but “why are you so bad at building sandboxes.”

As for real threats, Cory gives points to the plague of ransomware attacks mostly launched by “script kiddies” as an example of technology spread whether from leaks or by deliberate malice, to crooks and psychopaths. These “dumb-dumbs” do not need any special abilities to run the software. I’m more worried about these than by AI gone rogue.

Digital soverenty risks

Cory discusses how access to data and technology, or the lack of access, can be a powerful weapon wielded by bad actors, such as Vladamir Putin and Donald Trump. The most disturbing example is when Microsoft abetted Trump’s vendetta against the chief officer of the International Criminal Court. Microsoft, obliged the president by cutting of his Microsoft account, thereby disabling his email and other critical communications and productivity tools. (Note: Microsoft denies this, but the jury is still out. Anyway, it still serves as a chilling warning).

Cory points out that it would be possible to shut-off access to Microsoft 365 to an entire country, much like a neutron bomb it could devastate without wiping out the population.