Great sensing is wasted if the signal can’t travel.
Somewhere in your organization, right now, someone on the frontline has noticed something important. A shift in what customers are asking for. A competitor’s quiet move. A small failure that will become a big one. They’ve sensed the signal correctly. And it’s about to die in a manager’s inbox.
This is the failure that quietly decides competitive races, and it has a name: the transmission gap. The problem is almost never that the organization can’t sense — frontline people see early signals constantly. The problem is that the signal can’t travel. It gets stuck between the person who noticed and the people who could act, lost in a queue, softened in a status update, deprioritized by someone with no context for why it matters. Eighteen months later a competitor ships the very thing your frontline saw first, and everyone wonders how you missed it. You didn’t miss it. It just never arrived.
In structural terms, this is a broken feedback path — the link from the people doing the work to the people who run operations and set strategy — plus a missing or weak coordination layer. And the fix is not “sense more.” It’s to treat the organization’s signalling like infrastructure: to build, deliberately, the pathways along which intelligence travels — the organizational equivalent of a nervous system.
A nervous system has a feature worth copying directly: it doesn’t route everything the same way. Most signals travel the normal pathways at a normal pace. But a few — the something is badly wrong signals — get a fast lane that bypasses the entire chain of command. Beer called this the algedonic channel (from the Greek for pain and pleasure): the alarm wire that lets a sharp, urgent signal jump straight to the top without waiting for permission at each layer. Your hand leaves a hot stove before your brain finishes forming the sentence “that’s hot.” Organizations need the same reflex — a way for a critical signal to skip the hierarchy entirely.
This is also where AI enters the story in a concrete, non-hyped way. The hard part of always-on sensing has always been capacity: you can’t have humans manually monitoring and routing every weak signal across a large organization. AI makes that feasible for the first time — detecting patterns, flagging anomalies, and routing signals to the right place in real time. It’s the variety-handling machinery Beer imagined at Cybersyn but couldn’t fully build with telex machines. The infrastructure he sketched, the technology can now actually carry.
And here is the warning that the next essay will pick up, because it’s the hinge of the whole series: more sensing is not better regulation. A nervous system with no judgment behind it is just noise at scale. Pipe every signal everywhere and you don’t get a more responsive organization — you get one drowning in alerts, where the algedonic alarm rings so often that people stop hearing it. AI can make this worse, not better: it can amplify the wrong signals, flood decision-makers beyond what they can absorb, and automate a controlling, centralizing response that crushes exactly the local autonomy you were trying to protect. The technology that solves the transmission gap can manufacture a new pathology — paralysis by signal.
So the takeaway has two halves. Build the nervous system on purpose — designed pathways, plus a real fast lane for the signals that can’t wait. But build it knowing that carrying a signal is not the same as understanding it. The pipes are necessary. They’re not sufficient. What turns signal into good response is judgment — which is exactly where this series goes next, into the era of cheap everything.
Next week: when AI makes content and execution nearly free, the one thing that becomes scarce is judgment.
Part of the countdown to Metaphorum 2026 — “100 Years of Stafford Beer” (17–19 September 2026, Alliance Manchester Business School, Manchester), where I’m convening a half-day “Cybernetic Futures” workshop [workshop slot — TBC]. Programme, themes, and registration: conference2026.metaphorum.org. Abstract deadline: 17 July 2026.