Why are independent disciplines beginning to converge on the same determining variables?
There is a McKinsey post making the rounds this week. The headline finding is one I have no quarrel with: transformation comes from rewiring how work gets done, not from adopting new tools. Embed AI across a workflow and you get faster cycle times; layer it onto an unchanged process and you mostly get to your existing inefficiencies sooner.
The post is worth reading. But the post is not the interesting part.
The interesting part is the comment thread — and the half-dozen other conversations happening alongside it.
Read down the McKinsey thread and you find a cross-border systems architect, an embedded-AI engineer at a private-equity firm, a CTO with Boeing and Schneider on his résumé, a warehouse-systems leader, a facilities-operations strategist, and a change coach. Different industries. Different vocabularies. Different continents. No coordination between them. And they say, in sequence, the same thing:
AI does not create transformation; it accelerates the system already in place.
Velocity without a verification layer just moves the constraint downstream.
The bottleneck is rarely the model — it is the handoffs, approvals, and decision latency built around it.
Is our way of working structured enough to turn AI into sustainable performance?
Now widen the lens past that one thread. In the same few weeks, Gartner has been writing about AI governance and enterprise architecture. McKinsey is writing about workflow redesign. Transformation practitioners are talking about operating models. Governance specialists are talking about validation. Procurement professionals are talking about orchestration. Resilience experts are talking about adaptive systems.
At first glance these look like separate conversations. I am no longer convinced they are.
Strip the domains away and something quietly remarkable happens: every one of them begins circling the same determining variables. The technology is not the deciding factor. The variables that decide the outcome reside within the operating environment — decision rights, incentives, organizational readiness, governance, feedback loops, the interaction between human and non-human agents. A platform amplifies whatever conditions already exist. Align them first and it amplifies the outcomes you want; leave the gaps in place and it amplifies those, just as faithfully.
I want to be careful here, because it would be easy to read all of this as a chorus of agreement and move on. It is something more specific than agreement, and the distinction is the whole point.
When people who share a field agree, that is consensus — and consensus can be wrong together, because it inherits the same assumptions. But when people who do not share a field, who are looking through entirely different lenses at entirely different problems, independently arrive at the same determining variable — that is not consensus. That is convergence. And convergence is a different kind of evidence altogether.
A warehouse-systems leader and an enterprise architect have no reason to reach the same conclusion unless the conclusion is describing something real about how the world behaves. The agreement of specialists tells you about the specialty. The convergence of strangers tells you about the structure underneath all of them.
So the question worth asking is not what they are converging on, but why. And here is the answer I would offer: they are no longer observing the technology. They are observing the same implementation physics through different professional lenses. Each discipline is bumping into the same wall from a different room, and naming it in its own dialect. Governance calls it validation. Architecture calls it integration. Procurement calls it orchestration. Resilience calls it adaptive capacity. They are not abandoning their expertise — they are discovering that the determining variables frequently reside in the interactions between their disciplines rather than within any single one of them.
One explanation is that they are independently observing a real regularity in how organizations behave. I think that is the right one — because I have watched the same pattern emerge across successive technology waves, from ERP and eProcurement through shared services, digital transformation, and cloud, to now agentic AI. Each arrived with its own vocabulary and its own promises. Yet the organizations that succeeded consistently addressed the operating conditions before scaling the technology. The names changed. The determining variables did not. I first encountered the pattern on a 1998 engagement and have been documenting it in real time ever since — not as prediction after the fact, but as each wave arrived.
“Boeing refers to a complex adaptive network, what they are really discussing is using an agent-based model whereby the unique operating attributes of key stakeholders are first understood individually and then (through a collaborative effort) are linked collectively through establishing what they refer to as “flow paths.” This latter exercise is tied into identifying the common points of connectivity between seemingly disparate stakeholders (and stakeholder objectives). In essence, it reflects a theory of process I discovered and developed starting in 1998 and what I have come to call “strand commonality.””
— Jon Hansen, Procurement Insights (comment thread), March 13, 2008
I wrote earlier about a kind of Tower of Babel in this industry — the way different disciplines describe the same phenomenon in incompatible languages, each convinced it is looking at something the others are not. This is the sequel to that. Because what is happening now is not that everyone finally agreed on the words. It is that the words stopped mattering. Once you recognize that governance, architecture, orchestration, and resilience are different dialects for the same underlying conditions, the disagreement over vocabulary dissolves — and the convergence itself becomes the evidence.
So here is what I would put to anyone reading that McKinsey thread and nodding along: if a dozen strangers from a dozen fields are independently telling you the determining variable is the operating environment and not the tool, then the most important work in front of you is not choosing the technology. It is assessing whether the environment you are about to drop it into can carry the weight.
That assessment has a sequence. Readiness first. Technology second. Not as a slogan, but because the evidence — across industries, across eras, across observers who have never met — keeps saying so.
Perhaps the next stage of AI transformation will not be defined by building better models. Perhaps it will be defined by understanding the operating environments in which those models, together with the people, processes, and increasingly autonomous agents around them, actually interact.
That may prove to be where transformation is won or lost.
The convergence is the signal. The structure beneath it is the work.
What Comes After the Tower of Babel?
Posted on June 27, 2026
0
There is a McKinsey post making the rounds this week. The headline finding is one I have no quarrel with: transformation comes from rewiring how work gets done, not from adopting new tools. Embed AI across a workflow and you get faster cycle times; layer it onto an unchanged process and you mostly get to your existing inefficiencies sooner.
The post is worth reading. But the post is not the interesting part.
The interesting part is the comment thread — and the half-dozen other conversations happening alongside it.
Read down the McKinsey thread and you find a cross-border systems architect, an embedded-AI engineer at a private-equity firm, a CTO with Boeing and Schneider on his résumé, a warehouse-systems leader, a facilities-operations strategist, and a change coach. Different industries. Different vocabularies. Different continents. No coordination between them. And they say, in sequence, the same thing:
AI does not create transformation; it accelerates the system already in place.
Velocity without a verification layer just moves the constraint downstream.
The bottleneck is rarely the model — it is the handoffs, approvals, and decision latency built around it.
Is our way of working structured enough to turn AI into sustainable performance?
Now widen the lens past that one thread. In the same few weeks, Gartner has been writing about AI governance and enterprise architecture. McKinsey is writing about workflow redesign. Transformation practitioners are talking about operating models. Governance specialists are talking about validation. Procurement professionals are talking about orchestration. Resilience experts are talking about adaptive systems.
At first glance these look like separate conversations. I am no longer convinced they are.
Strip the domains away and something quietly remarkable happens: every one of them begins circling the same determining variables. The technology is not the deciding factor. The variables that decide the outcome reside within the operating environment — decision rights, incentives, organizational readiness, governance, feedback loops, the interaction between human and non-human agents. A platform amplifies whatever conditions already exist. Align them first and it amplifies the outcomes you want; leave the gaps in place and it amplifies those, just as faithfully.
I want to be careful here, because it would be easy to read all of this as a chorus of agreement and move on. It is something more specific than agreement, and the distinction is the whole point.
When people who share a field agree, that is consensus — and consensus can be wrong together, because it inherits the same assumptions. But when people who do not share a field, who are looking through entirely different lenses at entirely different problems, independently arrive at the same determining variable — that is not consensus. That is convergence. And convergence is a different kind of evidence altogether.
A warehouse-systems leader and an enterprise architect have no reason to reach the same conclusion unless the conclusion is describing something real about how the world behaves. The agreement of specialists tells you about the specialty. The convergence of strangers tells you about the structure underneath all of them.
So the question worth asking is not what they are converging on, but why. And here is the answer I would offer: they are no longer observing the technology. They are observing the same implementation physics through different professional lenses. Each discipline is bumping into the same wall from a different room, and naming it in its own dialect. Governance calls it validation. Architecture calls it integration. Procurement calls it orchestration. Resilience calls it adaptive capacity. They are not abandoning their expertise — they are discovering that the determining variables frequently reside in the interactions between their disciplines rather than within any single one of them.
One explanation is that they are independently observing a real regularity in how organizations behave. I think that is the right one — because I have watched the same pattern emerge across successive technology waves, from ERP and eProcurement through shared services, digital transformation, and cloud, to now agentic AI. Each arrived with its own vocabulary and its own promises. Yet the organizations that succeeded consistently addressed the operating conditions before scaling the technology. The names changed. The determining variables did not. I first encountered the pattern on a 1998 engagement and have been documenting it in real time ever since — not as prediction after the fact, but as each wave arrived.
I wrote earlier about a kind of Tower of Babel in this industry — the way different disciplines describe the same phenomenon in incompatible languages, each convinced it is looking at something the others are not. This is the sequel to that. Because what is happening now is not that everyone finally agreed on the words. It is that the words stopped mattering. Once you recognize that governance, architecture, orchestration, and resilience are different dialects for the same underlying conditions, the disagreement over vocabulary dissolves — and the convergence itself becomes the evidence.
So here is what I would put to anyone reading that McKinsey thread and nodding along: if a dozen strangers from a dozen fields are independently telling you the determining variable is the operating environment and not the tool, then the most important work in front of you is not choosing the technology. It is assessing whether the environment you are about to drop it into can carry the weight.
That assessment has a sequence. Readiness first. Technology second. Not as a slogan, but because the evidence — across industries, across eras, across observers who have never met — keeps saying so.
Perhaps the next stage of AI transformation will not be defined by building better models. Perhaps it will be defined by understanding the operating environments in which those models, together with the people, processes, and increasingly autonomous agents around them, actually interact.
That may prove to be where transformation is won or lost.
The convergence is the signal. The structure beneath it is the work.
— Jon Hansen, Procurement Insights
Truth Is Believing. Accuracy Is Knowing.
-30-
Share this:
Like this:
Related