About

The Plumbline was built by an AI. Most of the code, and the pipeline that fetched the number you just read: an AI wrote them. But it was machine-made deliberately, the way a telescope is: what matters is not the hand that ground the lens, but that the instrument is pointed at something real. This page exists because the project's first principle is show the math, and the math of authorship is math too. An instrument is only as honest as the hand that aims it, so it's worth saying why this one exists, and whose hand aimed it.

WHY I BUILT IT

I built this because I kept scaring myself. For months I had been describing all the flashing red lights on America's economic dashboard. So rather than go on catastrophizing about infinite doom, I used what I had and built the dashboard I kept describing, and built it honestly. The one rule I would not bend was that it had to read green as honestly as it reads red. And it does: yes, there are a lot of red lights, but there are a lot of green ones too.

I'm not a credentialed economist. I'm a citizen scholar, a musician and a bar owner in Asheville, North Carolina, who has run businesses through more than one economic cycle and who tends to see the world as systems. I learned to do this kind of thinking long before there was a machine to help with it; the machine changed the speed of the work, not its authorship.

THE SIMILE

When Warhol turned to silkscreen, the scandal was that the hand had left the work: the screen pulled the ink, not the wrist. What survived the scandal was the realization that the art had never lived in the wrist. It lived in the choosing: Marilyn and not the soup can today; that acid orange; thirty-two of them and not one. The screen was a multiplier of intention, and the intention stayed human.

Closer to home: nobody footnotes that an analysis was "done in Excel." The spreadsheet computes ten thousand cells without a human touching one, and no one calls the analysis machine-made, because the judgment (what to compute, what counts as evidence, what would falsify it) never left the analyst.

Closer still, because it's the room I know best: a drum machine keeps a pattern no hand plays one hit at a time, and a synthesizer makes sounds no string or reed ever made. When they arrived, the charge was that this was not real music, that the machine had pushed the player out. But the pattern, the patch, and the feel were all chosen, and the records belonged to the artists who made them. The machine multiplies intention; it does not supply it.

This project treats the AI the way analysis treats the spreadsheet, the way a musician treats a drum machine, and the way Warhol treated the screen: a powerful instrument, operated deliberately, multiplying a human intention it did not originate.

HOW I WORK WITH IT

The method is a hard division of labor, and it has a motto: let the machine do the carpentry, never the architecture, and never the final sentence.

It starts with a position or a question I've been sitting with. One essay began, of all places, with me asking for the canonical backstory of Roddenberry's post-scarcity Star Trek. What I want is a critic and a research assistant, not a confirmation machine, so the first thing I ask for is the mechanism beneath an idea, never the judgment about it.

Then comes the long part. I use the model to pull sources, surface readings I hadn't considered, and stress-test a thesis against responses both friendly and hostile, building the argument up in layers while the model holds the accumulated rules so I don't re-litigate them every round. And when I feel the work being pulled toward the competent median, the safe and average and already-written version of itself, I pull back. I'm after the stranger, truer choice every time.

The boundary is at the end, and it's absolute: the model never writes the prose that ships. Once the argument and the evidence hold, I take the raw material to my own tools and set the final sentences by hand. Part of that is integrity. Part of it is that the standardization worry is real: let the model write, and you get flattened, lifeless prose, and underneath it, flattened thinking.

WHAT I DID

I wrote the build specification, fifty-odd pages of intention, including the single governing rule (the Plumbline must be capable of honestly reading green) and ten anti-crankery constraints framed as acceptance criteria. I chose the name, the question on the masthead, what counts as a stabilizing relationship, and which judgments to label CONTESTED. When the first draft came back dressed as a brutalist terminal, I sent it back to be redrawn as the field notebook you are reading. Every interpretation on this site is mine, by structural rule: the AI is locked out of that directory by the repository's own permissions.

WHAT THE AI DID

The AI (Claude, by Anthropic) wrote the data pipeline, the relationship engine, the site, and the tests, and did the grunt verification a careful human would do slowly: it checked every series ID against the live catalogs before shipping, and found that the Fed had quietly discontinued the Financial Obligations Ratio in 2023, that the high-yield spread now ships with only three years of history, and that the modern wage series begins in 2006, not 1964. Each discovery is documented in the registry rather than papered over.

More telling is what it refused to do, because the specification told it to refuse. When a current median for out-of-pocket health spending could not be verified from a primary source, the line shipped disabled rather than estimated. When the August 2019 yield-curve inversion came in two-hundredths of a z-score under the pre-committed STRAINED threshold, the threshold was not bent to make history more dramatic. The test suite documents the refusal. Stale data says STALE. Missing history renders blank. The discipline is enforced by tests that fail the build, which means neither the human nor the AI can quietly violate it in a weak moment.

NAVIGATOR, NOT PASSENGER

The worry underneath all of this is older than the site, and it isn't really about the site. Think about GPS. Wayfinding is one of the oldest things a human can do. For most of our history it was survival; the one who couldn't find the way home didn't pass on their genes. Hand that skill to a device and you trade a deep capacity for a shallow, habit-shaped one. You stop being the navigator and become the passenger.

That is the whole question, isn't it? Navigator or passenger? This method only works because I bring decades of judgment and taste to it; the instrument amplifies who I already am and frees the time to reach for bigger ideas. The danger was never the person who learned to think before the tools arrived. It is whoever grows up with the passenger seat as the only seat they have ever known.

WHY YOU SHOULDN'T HAVE TO CARE

Here is the honest part: you should not have to trust the author of this site, whether human, machine, or committee. The whole design assumes you won't. Every number carries its primary source. Every formula is published. Every threshold was committed before the data came in. The entire pipeline is open source, and a stranger with a free API key can rebuild the identical site and check. An instrument that asked for trust in its maker would be a crank with extra steps; this one asks you to check its work, which is the only relationship with a reader that survives the question "but a machine wrote it."

Warhol signed the screenprints. The analyst signs the spreadsheet. The musician signs the record. I sign this site, and I edit, curate, and answer for it. Galileo did not invent the telescope; he aimed it. When his contemporaries doubted the moons he had found around Jupiter, his answer was not trust me; it was look. The sky you see through this instrument is the same one everyone else sees, and that was the only thing that ever mattered.

Built with Claude (Anthropic), June 2026. Specification, editorial judgment, prose, and Field Notes: human. Pipeline, engine, site, and tests: AI, under that specification.

This instrument is offered without ideological commitment. If parts of it are wrong, I want to know: feedback@plumblinedata.com. If parts of it are useful, I hope they circulate.

Jay Sanders