Why Was Gunkel So Obsessed with His Charts?

Those who stick with me here on Idea Chemistry are going to be rewarded, as from time to time I will post original Gunkel material that has never been shared publicly before.

Today I’ve decided to share a massive chart called “Future Scenarios Involving Metamachines.”

Metamachines was Gunkel’s word—in our current parlance—for artificial general intelligence (AGI).

The chart actually deserves its own post to analyze in light of the current Generative AI revolution, but I will have to defer on that to next year probably.

As you can see, the chart visually demonstrates, via individual scenarios like “Metamachine Runs for President” and “A.I. Transforming Art”, the connections between ideas under a single unified concept.

Some of these scenarios were written about by Gunkel elsewhere, such as his unfinished manuscript More Human Than Man.

Some of these scenarios—like “A.I. Transforming Art”—are already coming about.

In all, Gunkel identified well over two hundred future scenarios involving AGI on this chart alone, and probably many more scenarios in his writings.

What I want to pivot to here, however, is the theme I started in a prior post, which is that Gunkel’s charts are basically visual lists.

They give some insight into what was going on in Gunkel’s eidetic brain.

And as the student of ideonomy will eventually realize, lists are crucial to the empirical method Gunkel developed for a science of ideas.

Gunkel wasn’t always interested in visualizing his ideas.

In fact, it doesn’t seem that he was aware of the potential until his late 20s or early 30s.

Gunkel always attributed his experience with the Hudson Institute and his friendship with Herman Kahn as the source of inspiration for his “charts,” but I haven’t been able to make an exact connection between Gunkel’s charts and the way the Hudson Institute represented data in the so-called “chartbooks” it published.

For comparison, here’s an example of one page from a Hudson Institute “chartbook” from 1973. This is the only example of a chartbook I could find online without exerting an undue amount of effort. The chartbook served as a sort of appendix to the Institute’s policy analysis of increased legal gambling in New York. And as you will see, the chartbook doesn’t have a whole lot in common with Gunkel’s own chart.

Was Gunkel simply thinking about “charts” as a sort of shorthand for “large visual aids”?

More research would be needed.

From Sharon Gamari-Tabrizi’s book The Worlds of Herman Kahn, it seems that Kahn would accompany his very popular presentations on thermonuclear war and other such Strangelovian topics with a variety of visual aids.

One description of a mid-1950s Kahn briefing described by Gamari-Tabrizi includes a “young man” jumping onstage to help Kahn “with an armload of charts.”¹

Was this, at any time, Gunkel himself? The personal assistant to the great Herman Kahn?

More research would be required to find out.

Gunkel did use large poster boards to display chartbook-like work.

However, these works were much less common.

One of the strangest, but also most similar, examples of a Hudson-Institute style chartbook can be found on MIT’s ideonomy website in Gunkel’s analysis of visual structure.

This chart shows (I think) the way Gunkel tried to understand visual structure as being composed of certain universal forms. And he tried to plot this out mathematically using a technique known as multidimensional scaling.

To my knowledge, Gunkel did not specifically provide a formal name for “charts” in the scheme of ideonomy, though it would be very interesting for someone to search through his writings to find out whether he developed any technical terms or taxa.

So we have established the concept that “charts are visualized lists,” with the title of the chart (usually in the center) being the title of the list, and the list items being arrayed around the center.

In some cases, though, Gunkel took this idea one step further and turned his charts into ontologies.

What’s an ontology?

It’s basically a structure of knowledge that takes the idea of a list and expands on it.

I don’t know if anyone has ever come up with a “theory of lists,” but I guess you could say an ontology transforms a two-dimensional structure into a three-dimensional one.

As shown in the Gunkel chart below, “Things For Which Ocean Is A Metaphor” (posted on MIT’s website), beyond simply providing the chart’s title and content, Gunkel has actually used colors to link common themes together within the chart itself.

Natural terms are highlighted or circled in green—words like soil, atmosphere, galaxy, life, and storm.

Thematic aspects that are similar to the divisions of ideonomy (more on this later) are circled with orange: the Impossible, History, Future, Chance, Chaos, Beauty, Order, and Possibilities, for example.

This approach, if applied rigorously and consistently—which it doesn’t seem that Gunkel did—will ultimately create a dense network of interrelationships within the context of the ideas provided.

Circling back to the “Future Scenarios Involving Metamachines” chart, we can recognize that Gunkel actually used different colors to enclose different scenarios.

Were these just random decisions he was making, or was there a scheme behind his color-coding of the scenarios?

It’s just one of many things I constantly notice about Gunkel’s work that I don’t have any time to investigate further.

I’m telling you, there’s a massive new space here for Gunkel scholarship… it’s just waiting for people who are willing to take up the banner and give their papers an opportunity to pass peer review.

Sooner or later, something will get through.

And when the dam breaks….

(1) The Worlds of Herman Kahn by Sharon Ghamari-Tabrizi, p. 12.