I Owe Frankenstein an Apology
In which the author of a framework about observer effects discovers he is an observer, makes the exact error the framework predicts, and has to eat his own cooking in public.
In my previous piece, Spot the Template, I identified four template families operating in modern political communication. Protocols. Institutional. Dispensationalist. And Frankenstein — the thing we created that turns on its creator. 207 years of conserved momentum since Shelley. AI panic, nuclear fear, genetic engineering horror. I was very pleased with this. It was neat. It was clean. It explained a lot.
It was wrong.
Not wrong in the way you might think. The observations were correct. AI panic is real. Nuclear fear is real. The pattern of “we made a thing and now we’re terrified of the thing” is genuinely recurring across centuries of human history. I wasn’t hallucinating the data. I was misidentifying the structure.
Here is what happened, and I need to explain it carefully because the way I got it wrong is more interesting than the fact that I got it wrong.
The Progress Wave
After writing Spot the Template, I kept working on the formal framework. One of the things that emerged — and I mean genuinely emerged, I did not plan this — was the Progress wave. This is not a template. It is a property of the field itself. A dimension of the continuum. Think of it as the background frequency of communication technology transitions: stone to bronze to iron to writing to printing to radio to television to internet to whatever comes next. Each transition compresses the wavelength. The wave has been running since humans have been human. It is not an event. It is the field. I even discussed it here in the bread series:
The wave does something specific. It modulates every template it passes through. When a new compression event hits — a new technology, a new perturbation — each template produces its own response:
The Protocols template says: “They built this thing to control us.” Hidden hand. Coordinated threat. Named individuals behind the curtain. Big Tech oligarchs suppressing discussion of the dangers. You have seen this response to AI. You are seeing it right now.
The Dispensationalist template says: “This is the end.” The singularity. Superintelligence. The last invention humanity will ever need to make. p(doom). The timeline is running out. Secular eschatology with the serial numbers filed off.
The Institutional template says: “This needs a governance framework.” Safety boards. The EU AI Act. Alignment protocols. Senate hearings. Responsible AI charters. Procedural order converting the perturbation into a process problem.
Three different outputs. Structurally unrelated. Produced by three different templates in response to the same wave perturbation.
And from my position on the wave, looking at these three outputs, I did exactly what my own framework says every observer does.
I pattern-matched them into a single thing and called it a template.
The Error
I was sitting in a positive position on the Progress wave. From that position, all three negative modulations — all three fear responses to the new technology — looked like they were pointing the same direction. Away from me. Toward fear. I saw the apparent unity and I named it “Frankenstein.”
This is precisely the observer-position error that the Pekka Principle starts to discuss. The observer’s position determines what they see. An observer sitting at a positive position on the progress wave sees all negative modulations as convergent. They look like one thing. They are not one thing. They are three separate things that happen to produce similar-looking outputs from the observer’s vantage point.
I made the error. My own framework predicts the error. I did not see it until I formalised the Progress Wave as a field property, at which point the false template dissolved like morning fog.
I would like to emphasise that I wrote the framework. I wrote the section about observer effects. I wrote the bit about how your position determines your measurement. And then I went and made the exact error I had just described, because information is local and I am inside the field just like everyone else.
What Frankenstein Actually Is
“Frankenstein” is not a template. It is a wave artefact. It is what happens when the Progress wave hits all three templates simultaneously and an observer who cannot see the wave pattern-matches the three outputs into a false unity.
Every technology panic in history is this same dissolution. The printing press terrified the Church (Dispensationalist: heresy will spread, the end times), terrified the state (Protocols: seditious material, coordinated rebellion), and generated regulatory responses (Institutional: licensing, censorship, the Stationers’ Company). Three responses. Not one. The observer in 1500 saw “fear of printing” as a single phenomenon. It was not. It was the Progress wave modulating three templates.
Nuclear fear. Same structure. Protocols: “They have the bomb and they’re going to use it on us.” Dispensationalist: “This is Armageddon.” Institutional: “We need arms control treaties.” Three outputs. One apparent phenomenon. Wave artefact.
AI. Same structure. Same wave. Shorter wavelength. I listed the three responses above. They are not a debate. They are three monologues happening in the same room, and observers without the framework think they are hearing one conversation.
The Happisburgh Hand Axe
So what is AI, if it is not Frankenstein’s monster?
It is a tool. It sits inside the Progress wave compression sequence alongside every previous tool. Flint knapping at Happisburgh, 950,000 years ago. Bronze. Iron. Writing. Printing. Radio. Television. The internet. Each one compressed the wavelength. Each one generated the same three-template panic response. Each time, observers experienced the new tool as categorically different from all previous tools. Each time, they were wrong.
The equations do not need modification. The regime changes. The dynamics are the same. The only variable that changes is the wavelength.
A large language model is a Happisburgh hand axe. Just a bit more complicated.
And yes, I used one to help me work this out. I am aware of the irony. The hand axe helped the toolmaker see that it was a hand axe. The framework applies to the tool that helped build the framework. Information is local. Even for AI systems. Even for frameworks about AI systems. Even for authors of frameworks about AI systems who got their own framework wrong and are now correcting themselves in public.
The Apology
I owe the framework an apology for not trusting it. The framework said: there are three template families, and they derive from three cognitive primitives (Bartlett, 1932 — agency detection, temporal narrative, procedural order). The framework said: a fourth template would require a fourth cognitively fundamental schema. The framework said: apparent unities that span multiple templates are probably wave artefacts.
I heard all of this. I wrote all of this. And then I saw AI panic, nuclear fear, and genetic engineering horror all pointing the same direction and I said “ah yes, a fourth template.”
The framework was right. I was wrong. The framework caught me. It took a while, but it caught me.
I also owe Frankenstein an apology. Mary Shelley’s novel is not a template. It is a work of art produced by someone sitting on the Progress wave in 1818, observing the industrial revolution’s modulation of her own template position, and writing the output. The novel is the wave artefact, crystallised in prose. It is beautiful. It is not a political communication structure. I was wrong to classify it as one. Sorry, Mary.
The Dual Use Note
Here is the uncomfortable bit. This correction is itself a demonstration of the dual use problem I’ve now added to the framework’s introduction.
By publishing this correction, I am performing a measurement event. My position in the field changes. Your position changes (you now know the framework self-corrects, which changes how you evaluate it). The correction generates its own gravitational effects — it is, right now, attracting or repelling readers based on their template position. Someone in the Institutional template is thinking “good, he follows proper methodology.” Someone in the Protocols template is thinking “he’s covering his tracks.” Someone in the Dispensationalist template has already stopped reading.
I can see this happening because I understand the framework. But understanding the framework is what caused the problem in the first place — I saw the wave artefact and named it because I had the vocabulary to name things. The vocabulary is a tool. The tool can build or destroy. The dual use problem is irreducible.
All I can do is be honest about the errors and publish the corrections. The framework does not leave the observer unchanged. It did not leave me unchanged. I am not the same person who wrote Spot the Template. That person had four templates. I have three, and a Progress wave, and a much better understanding of why I keep getting things wrong from exactly the position the framework says I will get things wrong from.
Information is local. Even for the person who wrote the equation.
The framework described here is developed formally in The Rent Theory of Political Identity, currently under peer review, and in working papers on field-theoretic models for political identity capital. The book will be called The Outrage Dividend.
All proceeds from The Angry Dogs are donated to Ukrainian causes.