The thesis of the sociotechnical bridge

A system can work flawlessly and still be of no use to anyone. It happens all the time: the assistant answers, the dashboard refreshes, the digital procedure is tidier than the paper one, and people go on doing things as before. This chapter explains why that happens and what it takes for it to stop happening. It is the book’s central argument, and if only one chapter were read, it should be this one: the method, the concepts and the applications that come later are its unfolding.

The thesis in a sentence

Every organization is, at the same time, a technical machine and a fabric of people. That is why no technology truly transforms it unless one first learns to read the human side that holds it up. Whoever develops that double view, and above all whoever learns to build on both planes at once, ends up in an uncommon place. Not by luck, but because almost no one trains to inhabit both dimensions.

What follows adds no new ideas to that conviction. It aims at something more modest and more useful: to put it to the test, give it order and bring it down to concrete work.

The problem it names

Organizations arrive at artificial intelligence from very different starting points. They form a spectrum rather than a single situation, but they all end up in the same place.

At one extreme is apparent scarcity: the organization that believes it has nothing. Pure zero, in fact, hardly exists anymore. There are spreadsheets, a billing system, the chat where the real work is coordinated. What is missing is not technology but design, because those systems grew on their own and live in people’s heads. The task is to build a first backbone, and the temptation is to treat that construction as a purchase.

In the middle are the islands: systems accumulated over the years that do not talk to one another, with each area owning its own version of the truth. The data exists, but fragmented; the task is no longer to have it, but to integrate it.

At the other extreme is the mature but fragmented organization. It already has solid systems and now needs to orchestrate them and take the leap: add AI where it adds judgment and govern it with confidence. Here the obstacle is not scarcity, but inertia.

Across the whole spectrum runs the case that repeats most often: the project that did not pay off. A considerable share of digital initiatives does not deliver the value it promised, and the technology is rarely to blame. What usually happened is that it was installed on top of an organization no one had read. A broken process was automated, a redundant procedure was digitized, AI was handed to a team that does not trust it, a dashboard was built that no one looks at because it measures what is easy rather than what matters.

In every case the underlying error is the same: treating a sociotechnical problem as if it were a technical one. The question that decides the outcome is not “which model do we use?”. It is “what do these people do, why do they do it this way, what do they gain and what do they lose if this changes, and who will resist?”. Without that question, the best engineering available amplifies the disorder instead of correcting it: automating a broken process only lets you make the same mistakes faster.

An episode of the kind that keeps recurring in the field shows this clearly. Before it asked for the churn model told in The four links, Distribuidora Norte, a fictional food-and-beverage SME, bought an assistant that answered questions about customers from its database. Technically, the system worked. Two months later, no one was using it. The reason was not in the model. The database was filled in by three teams with different criteria: one used the “status” field for collections, another for shipping, and the third left it empty. So the assistant answered with data the staff themselves knew to be unreliable. The project failed because it was built on information whose meaning depended on a human process no one had read, and no improvement to the model would have saved it.

The same root shows up on different ground. A city government digitizes a procedure that used to be handled at the counter. The system is tidy, fast and well built. Six months later, the procedure still comes in on paper. The employees who handled it found a thousand reasonable motives to keep the old channel alive: the elderly resident who cannot manage the screen, the exception the form does not contemplate. No one sabotaged anything openly. The system simply took away from one area the control of a process that was, without appearing on any org chart, its source of power and part of its reason for being. It had been read as a software problem when it was, from end to end, a problem of who decides. The technology worked; what no one had looked at was whom it displaced.

The other axis: the adoption ladder

The spectrum says where an organization starts from. The other axis says where it leaps to. Adoption is not a switch that turns on all at once, but a ladder climbed step by step. At the bottom is everyday office work. One step up are the systems that order the operation: an ERP that integrates administration, a CRM that gathers the customer relationship. Higher up is predictive AI, which uses accumulated data to anticipate rather than merely record the past. And on today’s step is the agentic layer: software that no longer only suggests, but executes tasks end to end.

For the thesis, what matters is not the inventory of steps but their slope: the higher the step, the dearer the cost of skipping the human reading. In office work, a badly read system is simply avoided, and people go back to their spreadsheet. In the agentic layer, a process no one understood does not run more slowly: it runs on its own, at scale and with no witness. Picture an agent that approves refunds, routes complaints or prioritizes case files according to a rule no one stopped to audit. The error stops being a misfiled paper and becomes an automatic decision, repeated thousands of times before anyone looks at it.

Technical sophistication, then, does not soften the sociotechnical demand; it multiplies it. That is why the thesis does not age with the technology: each new step makes it more urgent. The same ladder is named precisely in Concepts of our own and becomes measurable, with levels 0 to 5, in IMIA; here it is enough to grasp the slope.

Why it is true (and not a comfortable opinion)

The thesis is not a consulting intuition. It is an empirical finding with seventy years of backing, sustained by disciplines that do not talk to one another and that nevertheless reach the same conclusion.

The origin lies in the Tavistock Institute’s studies of the mechanization of English coal mining. The new technology, the longwall method, was objectively superior; even so, productivity fell and absenteeism rose. The reason was not in the machine, but in the fact that mechanization broke apart the social organization of the work. Before, the miners worked in small, self-regulating teams that shared out tasks, covered absences and kept morale up underground. The new method split them into specialized, interdependent shifts, and put nothing back to fulfill that function.¹ The lesson founded a discipline: the technical and the social subsystem are not optimized separately; they are optimized together or fail together. Improving only one, however real the improvement, can degrade the whole.

The later tradition of sociotechnical design turned that principle into method. A design that does not involve those who will use the system produces systems that people sabotage, avoid or misuse. That is why user participation is neither courtesy nor a search for consensus. It is requirements engineering, because only users know the real process the system will have to support.²

The sociology of technology adds the mechanism that explains why this happens. An artifact does not “work” by virtue of its internal properties, but when it manages to translate the interests of the actors around it: to enroll them and align them enough that each one recognizes themselves in it and finds a reason of their own to sustain it. Where that translation fails, the artifact is neither good nor bad; it is simply ignored.³ Adoption, then, is neither a step that comes after design nor a training problem to be solved later. It is the test of whether the design understood anyone.

In the age of AI the pattern does not disappear; it gets worse. Human-centered AI research shows that systems that isolate, obscure or replace the person generate rejection and harm, while those that amplify the person’s judgment, give them control and earn their trust, pay off.⁴ The field evidence points the same way. In an experiment with hundreds of entrepreneurs, generative AI helped the high performers and harmed the low performers, because the latter followed generic advice without the judgment needed to filter it.⁵ Without a human reading to guide it, AI does not level the field: it widens the gap that already existed.

Four perspectives, one conclusion. The industrial sociology of the 1950s, the tradition of participatory design, the sociology of technology and contemporary AI research start from different objects: a coal mine, an information system, any artifact, a generative model. Decades separate them and, even so, they converge on the same point: the performance of a technology is not decided by its technical merit, but by how it meshes with the human fabric that receives it. That such distant disciplines arrive at the same place is what makes this a finding rather than an opinion. The question is no longer whether the human dimension matters, but who is in a position to read it.

The structural gap

If the problem is sociotechnical, the solution requires reading the human and building the technical. But the labor market is split into two halves that rarely touch.

On one side are those who understand people: those who know about processes, change management and organization, but do not build the system. They diagnose, deliver a careful report and withdraw; afterwards, the code is written by someone else who never read that report or sat down with the people it described. On the other side are those who build the system (developers, data engineers, architects), who rarely read the organization. They receive a requirement already translated, and badly, by a third party, and set about optimizing the wrong solution with all their craft. The talent on neither side is what fails. What fails is that they never met.

Between the two halves opens a translation gap, and that is where the project is lost. Each side blames the failure on the other (“the business doesn’t know what it wants” against “the techies don’t understand the problem”), and both are partly right. The cost of that gap is not abstract. It is paid in systems no one uses, in data no one trusts and, measurably, in direct loss: poor data quality costs most organizations a considerable fraction of their revenue.⁶ That liability is almost always the mark of a human process the technical side could not see and the organizational side could not correct.

An intermediary who coordinates the two parties does not close the gap. That is project management, and it leaves translation in the hands of a document that degrades at every step. What closes the gap is an integrated competence: the two readings exercised as a single practice, which reads the human fabric and builds the system without meaning being lost in the handover. That practice does not translate between two specialists; it embodies the translation.

The integration of two domains

Closing the gap requires bringing together two domains of competence that the market keeps apart:

• Reading the social subsystem. This is the ground of the sociology of organizations: power, culture, incentives, resistance, real processes as opposed to formal ones.
• Building the technical subsystem. This is the ground of software and data engineering: auditable systems, pipelines, infrastructure, and the design and governance of AI under the auditability demands of a regulated environment.

Most professional profiles were trained in one of the two domains and practice it competently within its limits. The paradigm calls for crossing that limit and practicing both domains as a continuum: diagnosis of the organization, data modeling, system construction and AI governance. Each stretch is tied to the previous one and answers for what that stretch left behind. That continuity, and not the sum of specialties, is what the method organizes in The four links.

Honesty criterion. The integration of competences is a structural and verifiable differential, not a proclamation: it rests on evidence —systems in production, MCP servers, the IMIA maturity model— and not on adjectives. Where a piece of data is an assumption, it is said. The paradigm is earned in the facts.

Implications

From the thesis follows, with no additional steps, everything the rest of the book develops:

1. First you measure maturity, then you choose the technology. The sociotechnical diagnosis precedes any purchase; it does not accompany it. Skipping it is the most expensive way to start: you acquire the right tool for the wrong problem, or the right one for a problem the organization is not yet in a position to sustain. That is why an instrument is needed to measure that starting point before deciding. → IMIA — the maturity instrument
2. Data quality is a social problem before it is a technical one. A “dirty” piece of data is rarely a data-entry error. It is usually the faithful mark of a badly designed or badly incentivized human process, like the three teams that fill the same field with different criteria because each uses it for something else. It is not cleaned with a script: it is corrected by intervening in the process that dirties it. → Concepts of our own
3. AI governance is not a brake but the condition of adoption. Auditability, traceability and human control are not opposed to the technology being used: they are what makes people trust it enough to use it. A system that cannot be explained is not adopted. It is tolerated, and abandoned at the first doubt.
4. Value is measured in the person, not in the model. The deliverable is never a model in production or a dashboard switched on. It is a better human decision, an hour freed from a task that did not deserve it, a gap that closes. When the metric shifts from the artifact to the person, it becomes impossible to confuse activity with result, and almost every project that does not pay off lives in that confusion.
5. AI is a multiplier with a sign, and the bridge decides the sign. The same tool amplifies whatever it finds. On order and judgment, it multiplies value; on disorder and without judgment to filter it, it multiplies harm and widens the gap instead of closing it. That is what the Kenya experiment cited above showed. The human reading, then, is what inverts the sign of the multiplier: it turns an amplifier of inequalities into a tool for closing gaps.

Scope and limits

A thesis that takes itself seriously must also say where it ends. Four clarifications mark that edge before the argument closes.

Thesis and practice: two registers

The sociotechnical thesis, valid regardless of who applies it, must be told apart from the integrated practice this book proposes: integrated competence exercised as a single journey. A well-led multidisciplinary team can embody the thesis. The one-person practice is a differentiated bet, scarce by structure, and not a law of the field.

When this approach does not apply

The bridge is not a universal answer. It does not apply when all that is sought is a turnkey tool, with no organizational change; when the problem is purely technical; when there is no openness to having the organization read; or when an honest diagnosis concludes not to do AI yet. In that last case, the value lies in a readiness report, not in a project.

Scalability

Artisanal continuity does not scale linearly. To grow, it combines a core of diagnosis and design, the supervision of internal teams, and instruments (IMIA, protocols, living documentation) that raise the floor without removing the ceiling of judgment.

Adoption without teleology

The ladder and maturity measured with IMIA provide a direction and a snapshot, and both are indispensable. But in SMEs adoption is often recursive and episodic, not an irreversible march (Amanollahnejad et al., 2026).⁷ A level that drops is not always a failure: it may be episodic alignment that comes undone for lack of hygiene or backup leadership. That is why the bridge does not promise joint optimization as a final destination. It promises to manage recurrent misalignment without automating the error faster. Sometimes the friction that appears after implementation is a design signal (ambiguous roles, unreliable data), and the answer is to return to diagnosis, not to accelerate the model.

Antithesis: what the thesis argues against

Everything above is easier to grasp by contrast. The thesis takes issue with four comfortable positions:

• Against technological solutionism: the belief that the problem is solved by buying the tool.
• Against diagnosis without construction: the organizational analysis that never reaches the system.
• Against technique without reading: the construction that does not understand whom it serves.
• Against AI enthusiasm: models adopted for fashion, measured by vanity metrics and not by value.

The position of the bridge is what remains once those four comforts are discarded.

See also: The four links · Concepts of our own · Authors and currents · Bibliography