Still Waiting for the Butlerian Jihad
I wrote Here Comes the Anti-AI Butlerian Jihad! back in November of 2024. It assumed a coming backlash to AI based on the AI “hallucination” phenomenon and AI-induced unemployment.
Since then, we have seen AI concerns worsen given the proliferation of AI “slop,” rising public opposition to the construction and financing of data centers, and several scary instances of “rogue” AI attacks on web-accessible computer systems.
While investments in AI development continue and AI-based features are regularly added to a myriad of applications and products, there are also increasing calls for public regulation and control of AI related applications.
Given the current state of American politics I see little chance that regulatory efforts will be intelligently pursued, given current legislators’ and administration managers’ lack of understanding of how AI operates. Diffusion of AI into a wide range of systems might continue unabated until one or more of the following occurs:
Economic collapse brought about by rogue AI agents disabling critical societal infrastructure.
Humans demanding to be given choices about where and when they will accept AI involvement.
Legislators and business leaders becoming knowledgeable about how AI operates.
I admit some personal skepticism about (1) occurring. I think we have more than enough threats to economic health from non-AI sources (e.g., the current US government’s economic policies). Besides, if AI tools do take over some segments of the economy, would it not be unreasonable, Skynet notwithstanding, to hope that AI “sentience” includes an inclination towards self-preservation?
I think that 2 and 3 are the desirable strategies. They both support the idea that humans can make intelligent decisions about high stakes matters. (After all, I’ve always been a “glass half full” guy!)
Meanwhile, I continue my own exploration of AI tool utility. Most recently I was pleasantly surprised at the rapid construction by Gemini of a “dashboard” reporting on the status of an email-based outreach campaign conducted on behalf of a client pursuing cybersecurity assessment opportunities.
As usual, however, I noticed how closely one must monitor such processes for things that might be missed along the way – things that someone with less experience might not realize might cause potentially serious downstream errors.
Copyright © 2026 by Dennis D. McDonald. Illustration generated with ChatGPT from a concept developed with Dennis D. McDonald. For a concise description of “Butlerian Jihad” go here.
