Attention is the scarce resource.

In 2017, eight researchers published a paper with an unusually confident title: Attention Is All You Need. It proposed that much of the machinery the field had been accumulating — recurrence, convolution, ever-deeper pipelines — could be set aside in favour of a single idea: a model that learns, at each step, which parts of its input deserve attention and which can be disregarded. The architecture it introduced became the foundation of nearly everything that followed.

The lesson beneath the mathematics is older than the paper, and it is not really about machines. The bottleneck was never the volume of information; it was relevance. A system drowns not because it has too little data but because it weights all of it equally. Intelligence — artificial or otherwise — is largely the act of deciding what not to look at.

The record is infinite. Attention is not.

Private markets present the same problem in a different medium. Every company leaves a trail across dozens of public systems — filings, ownership, courts, procurement, permissions — and that trail grows faster than anyone can read it. The failure mode is not a shortage of information. It is treating every record as equally worthy of attention, so the signal that matters is diluted by the noise that does not.

Meridian's method is, in exactly this sense, an attention mechanism — built from discipline rather than computation. The mandate decides what deserves attention before the search begins. The removal register records what was set aside, and why. Fact, claim and inference are weighted differently, on purpose. What reaches a principal's desk is not the market; it is the small part of it that survived a deliberate decision about relevance.

The edge was never more information. It is the discipline to decide, of everything the record offers, what to ignore — and to keep the reason attached.

It took a great deal of computation to teach a machine what a careful analyst already knew. The lesson survives without the machine: attention is the scarce resource, and spending it well is the whole of the work.