The AI Telephone Game

My first blog post generated by Claude AI based on my source document, my comments on some new Dynaxiom entries. I hadn't always been explaining them because aphorisms are about the reader's interpretation, but in such a low-context society the clarifications might be necessary. The video podcast was generated by Google NotebookLM based on the same source document. 

The Dynaxioms have seldom been grouped by topic. Sometimes I cross-referenced them, but rarely. But that's all that LLMs live for--sometimes interesting, sometimes not, but not to the point of confabulation. If it does make an error usually you can trace it back to something, such as a missing author name, in which case it would reference the only name it had in the source materials. On one occasion I had quoted Elon Musk but his name didn't appear in that particular data set, and it attributed it to me. 

***

The AI-generated blog post (with my edits)

There's a peculiar irony to our information age: the internet simultaneously reveals everything and hides just as much. We swim in an ocean of data, yet so much remains invisible until the moment we need it, want it, or stumble upon it. Like dark matter in the universe, this informational dark matter exists in a state of potential—inert until it becomes visible, until we decide it matters.

But what is visible isn't always what's important or interesting to us as individuals. It's just what's popular. As a civilization, we have become remarkably good at constructing realities around what captures our attention rather than what sustains us.

Consider how technology reshapes not just what we do, but how we think about doing it. In the late 1990s, Photoshop—software designed as a digital darkroom—inspired me to pick up a camera. The tool preceded the craft. The simulation sparked interest in the real thing.

We're witnessing something similar with AI today. Large language models aren't just tools for productivity; they're access points to other things. Someone playing with AI music generation might suddenly want to pick up guitar. 

New technologies don't just solve problems—they create new highways, new territories to explore. Whether we choose to travel them is another matter. Some of us become technology or tool collectors, but only scratch the surface of all of them, as opposed to taking a few of then (or only 1) and finding many things to do with it, or the opposite, taking the tool, like the smartphone, that does many things, and never have the time to explore its capabilities. When digital photography emerged; others never saw the need to upgrade from their point-and-shoot. Both approaches are valid, but they yield vastly different perspectives on the world. Someone today might only know the smartphone camera, one device that does everything adequately. But someone who's been photographing for decades, has used SLRs, medium format cameras, drones, instant cameras, knows how to choose the appropriate tool.   

The Copy-Paste Civilization

After a generation of the internet, we live in a world built from things that have been copied, pasted, and edited. Truth itself has become something between a consensus and a metaphor.

You can take three sources, feed them to an AI, generate a summary, and build an argument from that summary. Without citations, the authority becomes questionable—yet if it sounds true, people accept it as true. We've lost the ear for discord, the instinct that tells us when information sounds wrong the way a musician knows when a note is off-pitch.

This matters more than we realize. When you define something—or redefine it—you make it easier to define or redefine everything else. Define what art is, and you've defined who artists are. Decide that certain work is "real" art and other work is not, and suddenly you've created entire categories of what deserves your attention and what gets relegated to the basement.

But we can always redefine things; We can revisit, reconsider, change our minds without ever finally making them up. A Warhol print can move from the basement to above the couch and back again. Our definitions don't have to be permanent installations.

Doubt & Certainty

Doubt requires continuous validation. Certainty doesn't. When you're certain, you can rest. When you doubt, you must constantly verify, question, reassess. In an age where information multiplies exponentially and truth becomes increasingly metaphorical, perhaps doubt is the more honest position—even if it's the more exhausting one. LLM: Doubt and Certainty 

The Long View

Technologies have always created new worlds that become increasingly populated over time. What matters is whether we remain curious enough to explore them with multiple lenses, skeptical enough to question our patterns, and humble enough to redefine what we thought we knew.

The photographer with many lenses doesn't always take better pictures than someone with a smartphone. But they have more choices about how to frame what they see. In an age of infinite information and constructed realities, maybe that's the skill that matters most: knowing we have choices about how we see, and being willing to change our lens.

[In art skill doesn't even need tools at the beginning. It starts with ideas. A tool may give you an idea as a happy accident, especially in the beginning when you're exploring it. Fully knowing tools can sometimes take that magic away. This is why I like alternate tunings because the tool you know (standard tuning) is keeping you within those constraints. There are still unlimited possibilities, but not as many as using different tunings.] 

From Doubt and Certainty: "...if people do not know how things are going to turn out, they have a sense of freedom. Yet, even when they know the astronomical odds against winning the lottery, people still buy tickets. Either they "feel lucky," which is a way of saying that there is a hidden cause at work, or they buy ticket numbers corresponding to their birthdays, which is also saying there is something mysterious at work."