Workforce transformation executive Kason Morris says organizations are redesigning work they have never accurately described. Enterprises are committing billions to AI adoption, while few can articulate how work creates value inside their own walls. “AI automates tasks, not jobs,” he says. The documented version of a job has always been an approximation, and the people performing it have corrected for what the documentation missed.
That correction is what automation removes. A system built from the documented process executes the description faithfully, including the portions that were never accurate. Morris, founder of Execution Visibility, has spent nearly two decades helping enterprise leaders understand how work gets done before technology is applied to it. The visibility gap he describes has existed in every organization for as long as organizations have kept records. What has changed is the tolerance for it.
Seeing the Work Before Redesigning It
The foundation of responsible AI adoption begins with a clear view of the tasks, decisions, handoffs, and judgment that generate value across the business. Without that view, redesign proceeds from assumption rather than evidence. Assumption produces a predictable pattern. Organizations automate the work that is easiest to document, which is rarely the work carrying the most undocumented weight. The two are difficult to separate precisely because one is well described and the other is not. When work becomes visible, Morris says, organizations can redesign with intention instead of assumption, which changes the quality of every decision that follows.
Distinguishing Value Creation From Capability Building
Once the work is visible, a second distinction becomes possible. Some work is well suited to automation. Other work is how people develop expertise, judgment, and leadership, and Morris locates the advantage in knowing which is which. The developmental function is the clearest example of something no process map records. A junior analyst building a model produces a deliverable and becomes someone who understands the business, and only the first outcome appears in any documentation.
Automating the task preserves the deliverable, while the second outcome ends silently. Developmental work does not identify itself as developmental; instead, it appears as a routine task performed by someone junior. This is the profile automation targets first and the least likely to survive a cost review.
Designing Systems Where Both Grow
Morris rejects the idea that leaders must choose between automating work and preserving it. The organizations he describes build systems where AI improves execution, while the remaining work continues to develop human judgment and organizational capability.
That outcome requires deliberate decisions about which work to hand over and which to protect, and those decisions cannot be made responsibly against a description known to be incomplete. The sequence is the substance of his argument. Understand what work creates, then redesign it. Reversing that order produces automation that performs faithfully against an inaccurate map, and the load-bearing elements only become apparent once the people who had been compensating for them are gone.
“The organizations that thrive won’t be the ones that adopt AI the fastest,” Morris says. “They’ll be the ones that understand what work creates before they redesign it.” To learn more, connect with Kason Morris on LinkedIn.
