Adapted from a book in progress on AI, crypto, and the machine-readable company.
Most companies are trying to deploy AI into operating environments they cannot describe with a straight face.
That is the problem.
They say they want copilots, agents, automation, and decision support. Fine. The tools are real. The models are real. The leverage is real.
But most companies are still running on a mix of meetings, side channels, tribal memory, half-maintained systems, and a few indispensable employees who know where the real process is buried.
Then leadership acts surprised when AI sounds polished and behaves inconsistently.
It should not be surprising.
Most companies do not have an AI problem. They have an illegibility problem.
The company is not readable enough for AI to be reliably useful.
That is what I mean by a machine-readable company.
Not a company run by robots.
Not a company reduced to workflow diagrams.
Not a giant management dashboard pretending to be strategy.
A machine-readable company is one where the important parts of work exist in forms that humans, systems, and AI can all interpret from the same underlying truth.
That standard is much higher than most firms realize.
What machine-readable actually means
A machine-readable company can answer basic operating questions from system state instead of institutional folklore.
What is in flight?
Who owns it?
What changed this week?
What is blocked?
Which customer commitment is at risk?
What decision was made, by whom, and why?
Which exception path was used?
Where is work happening outside the official process?
Most companies cannot answer those questions cleanly without calling a meeting.
That is not because the people are weak. It is because the operating system is fragmented.
Part of the truth lives in the CRM.
Part lives in tickets.
Part lives in Slack.
Part lives in meeting notes.
Part lives in spreadsheets that everyone claims are temporary.
Part lives in the heads of the five people everyone calls when something breaks.
And then somebody asks AI to summarize reality.
Reality compared to what?
If the company has three versions of the truth, the model will smooth, infer, and guess. It may do it elegantly. That is what makes the problem dangerous. AI does not just expose broken state. It can make broken state sound coherent.
A bad system used to fail noisily.
AI lets it fail persuasively.
Three layers have to become readable
1. Work state
Tasks, priorities, blockers, dependencies, deadlines, ownership, and status changes must be recorded consistently enough that another system can interpret them without a human translator.
2. Decision state
Important decisions have to survive the meeting where they were made. If the company cannot preserve judgment in durable form, both humans and machines are forced to rediscover the same context over and over.
3. Runtime truth
Operational reality has to connect back to ownership and action. Deployments, incidents, security exceptions, failed flows, and customer-impacting events cannot remain isolated from the operating layer.
That is where “machine-readable” stops sounding neat and starts sounding necessary.
Why this matters now
Because AI is moving from the edges of work toward execution.
It drafts the memo.
It routes the issue.
It summarizes the account.
It explains the incident.
It proposes priorities.
It helps write the code.
It shapes what managers see before they decide.
That means bad state no longer just creates bad reporting.
Bad state creates bad action.
This is why so many AI rollouts look good in demos and feel thin in production. The model is capable. The company is not queryable enough to support the use case at scale.
Everyone wants intelligence. Very few want the discipline intelligence requires.
What a leader should do next
Start with a blunt audit.
Where does important work still depend on oral tradition?
Where do exceptions get handled off-system?
Where does ownership blur?
Where do teams maintain shadow records because they do not trust the official one?
Which decisions disappear after the meeting that produced them?
What would an intelligent system need in order to answer “what is at risk this week?” with confidence?
Then make a few hard choices:
pick the canonical system for work state
pick the canonical place where decisions get preserved
pick the operational signals that must connect back to owners and actions
standardize the minimum vocabulary that lets teams interpret state the same way
stop tolerating ten unofficial truths
That is the beginning.
Not an AI strategy deck.
Not innovation theater.
Not a cloud of pilots hovering above broken workflows.
A readable operating system.
The real point
The machine-readable company matters because it is the condition that makes AI useful without making the business more brittle.
If AI is the great centralizer, machine-readable work is what keeps that centralization grounded in reality instead of polished fiction.
If crypto is the great decentralizer, machine-readable work is also how firms interact with decentralized systems without drowning in complexity.
That is the bridge.
The next question is the one most companies are still trying to avoid:
Once you can see the mess clearly, what do you do with it?
That is the subject of the next post: Do Not Automate the Mess.
