The Quiet Rewrite
Why SAP’s AI transition is not about jobs—but about logic
The Numbers Don’t Add Up
There is something quietly disorienting about the numbers. Thousands of roles are being phased out. Entire functions are declared obsolete. And yet, the total number of employees at SAP remains largely unchanged. On the surface, it looks like continuity. Stability, even. But beneath that surface, something else is happening.
This is not a reduction of work. It is a replacement of its internal logic. What is being reshaped is not the workforce itself, but the structure through which work exists. A form of what could be described as ghost turnover—where the visible organism remains intact, while its internal composition is quietly exchanged.
Roles do not simply disappear. They are rewritten. Functions do not vanish. They are absorbed into systems. And slowly, almost imperceptibly, the identity of the organization begins to shift.
From Tool to Architecture
What once defined enterprise software was its position in relation to human action.
Systems were built to support decisions, not to make them. They structured workflows, enforced rules and ensured consistency—but always within a framework defined elsewhere. The human remained the origin of intent.
That boundary is now dissolving.
With the emergence of AI-driven layers, including agents like SAP Joule, interaction itself begins to change. The system is no longer navigated—it is addressed. Intent replaces instruction. Outcome replaces process.
What follows from this shift is subtle but profound: the system does not just execute work—it begins to shape the conditions under which work exists.
“We are entering a phase where AI is no longer a feature of software, but its architecture. This means we are not just redefining jobs, but the underlying logic of how an enterprise operates.”
Christian Klein, CEO, SAP
The implication is not simply technological. It is morphological. Work is no longer something that flows through a system. It is something that is increasingly formed by it.
The Disappearing Middle
Every organization has, or had, an internal layer that rarely appears on org charts but defines how things actually function. A layer of translation.
Between strategy and execution, there has always been a group of roles responsible for turning ambiguity into structure. They interpret intent, align processes, adjust for exceptions and ensure that systems reflect the messy reality of human activity.
This layer is not disappearing. But it is thinning. And in that thinning, something critical emerges: an interpretation gap.
When AI systems begin to translate intent directly into execution, the intermediate step—the human act of interpretation—is compressed or bypassed. The process becomes faster, more efficient, more scalable.
But also less visible.
“If the translating layer disappears, so does the shared understanding of why a process works the way it does. You automate the norm—but the meaning behind it becomes harder to access.”
Senior Partner, Tara Consulting Group
What remains is a system that functions, but is increasingly difficult to explain. A system that produces outcomes without exposing its reasoning.
A system that begins to resemble a black box, not because it is opaque by design, but because its interpretive layer has been abstracted away.
Reskilling the Role
This has direct consequences for how roles evolve. The language of reskilling suggests continuity. It implies that workers move upward, gaining new competencies, adapting to new tools. And in many cases, they do.
But what is less visible is that the nature of the role itself is being redefined. The process manager becomes a system supervisor. The analyst becomes a validator of outputs. The coordinator becomes an observer of flows they no longer fully construct.
This is not a simple transition from manual to digital work. It is a shift from participation to oversight.
“We are moving from human-in-the-loop to human-on-the-loop. Systems increasingly talk to each other. The human role becomes one of intent and boundary-setting, rather than execution.”
Strategic Lead AI Infrastructure, Nokia
The distance between action and understanding grows. And with it, a new form of dependency emerges—not on tools, but on the systems that define how those tools operate.
The Enterprise as Infrastructure
What makes this transformation particularly significant is the position SAP occupies within the global economy. It is not merely a software provider. It is an infrastructural layer.
Its systems define how organizations:
- allocate resources
- structure supply chains
- manage financial flows
- interpret compliance
When such a system changes, the organizations built upon it do not simply adapt—they are reconfigured.
This is where the notion of algorithmic determinism begins to take shape. The logic embedded within the system is no longer optional. It becomes the default architecture through which economic activity is organized.
“Companies like SAP are no longer building tools. They are building environments in which decisions about human activity are made. The question is no longer whether the system works, but whether we still understand the system we are working within.”
Prof. dr. ir. Frank Baaijens, Technische Universiteit Eindhoven
The shift is subtle because it is distributed. There is no single moment of change—only a gradual realignment of processes, roles and expectations.
A workflow is shortened.
A decision is automated.
A layer is removed.
And over time, the system begins to operate according to a logic that was not explicitly designed at any single point—but emerges from the accumulation of changes.
The Quiet Rewrite
This is not an event, but a transition. Not a disruption, but a reconfiguration.
It does not announce itself. It does not demand attention. It simply proceeds – through updates, integrations, optimizations.
Until the moment arrives when the system no longer behaves as it once did. And by then, the rewrite is already complete.
What This Series Explores
This series begins from that observation.
Not to analyze SAP as a company, but to understand what happens when systems begin to define the structures they once merely supported.
In the articles that follow, we will explore:
- what it means when interfaces disappear and systems are addressed through intent
- how human oversight becomes thinner and more abstract
- how enterprise software shapes economic reality at scale
- and what this implies for Europe’s position in a system-defined world
Because the question is no longer how we use systems. But how systems, increasingly, use us to complete their logic.
This article is part of The System Rewrites Itself, a series exploring how AI is reshaping the logic of work, control and enterprise systems.
Illustration: DALL·E / OpenAI (concept & direction by Altair Media)
Caption: Rewriting the blueprint: as enterprise systems evolve, the structure of work shifts with them.
