Are You Optimizing Jobs That are Disappearing?
Efficiency Gains: Companies are automating roles but must consider what human contributions are essential as AI evolves.
Job Market Shift: A significant decline in entry-level positions threatens long-term leadership development opportunities for organizations.
AI Investment Flaws: Most organizations incorrectly prioritize automation over addressing essential human roles and skill development.
Strategic Ownership: AI transformation responsibility should lie with CHROs to better align human work and technology.
Talent Development: Emerging workforce models require a focus on multifaceted skills, not just traditional specialization.
Somewhere, right now, a company is making an investment to make a role more efficient. The role processes claims, or routes requests, or synthesizes reports, something of that nature. The team mapped the workflows, identified the friction, and built a plan. By the end of the quarter, that role will take 30% less time to do the same work.
By the end of the decade, the work will be done entirely by AI.
This is not a story about the success or failure of technology. It’s the ongoing story of the most significant leadership challenge of this century, and it is happening at scale across industries that believe they are being strategic about AI while they are, in fact, just being busy.
Keep Reading—and Keep Leading Smarter
Create a free account to finish this piece and join a community of forward-thinking leaders unlocking tools, playbooks, and insights for thriving in the age of AI.
Have an account? Log In
The question underneath almost every serious AI conversation right now is the one most organizations are avoiding, Not how do we use AI, but what do we actually need humans for?
Those spending millions on productivity optimization are not answering that question. They are deferring it, one efficiency gain at a time.
The Pipeline is Already Collapsing
Thirty-seven percent of organizations plan to replace early-career roles with AI, according to Korn Ferry. Another 66% have slowed entry-level hiring. If you are a recent graduate trying to understand why the job market feels structurally different from what your professors described, this is why.
Entry-level roles have historically functioned as the primary mechanism for organizations to grow their own talent. The analyst becomes the manager. The coordinator becomes the director. The junior employee who spends three years learning the business from the inside becomes the person who eventually leads it.
Eliminate the entry point and you do not just lose cheap labor. You lose the developmental pipeline that produces experienced leaders a decade from now.
IBM is moving in the opposite direction, tripling its entry-level hiring as competitors pull back. Whether that is a talent bet or a PR play remains to be seen. But one thing that is not up for debate is the simple fact that you cannot develop leaders you do not hire. IBM is positioning itself to have humans at the wheel well into the future.
The Wrong Question is Getting all the Budget
Most AI investment today follows a familiar logic: identify where humans spend time, automate as much of it as possible, measure the productivity gain, repeat.
The 70-20-10 framework that we’ve dug into previously is how some sophisticated organizations think about AI allocation. It reflects a different priority — 70% of investment going to people and process redesign, 20% to infrastructure, 10% to the algorithms themselves. Just one small caveat, almost no organization actually operates this way. Most have it inverted.
The result is organizations that are technically sophisticated and strategically adrift. They have better tools and no clearer answer to what the humans operating those tools are supposed to become.
The roles most vulnerable to automation are not the roles that are least valuable. They are often the roles that require the most routine judgment, or the kind of structured problem-solving that turns out to be exactly what machine learning is good at.
The roles that remain, or that will be created, require something different: the capacity to navigate ambiguity, make ethical decisions, manage relationships across competing interests, and exercise the kind of contextual judgment that does not reduce to a prompt.
There aren’t many job descriptions that fit this from what I’ve seen. There is no career ladder for it at present and there is almost no development budget pointed at it.
Who Owns This Problem
David Swanagon has made the argument that AI adoption has been handed to CIOs when it belongs with CHROs.
Adoption is completely different than deployment. The CIO has been assigned not only the design, test, and deployment of tools, but also the adoption of them. Adoption should be owned by the CHRO — because it deals with culture, trust, autonomy, skills.
The logic is sound. Technology implementation is a relatively small part of what makes AI transformation succeed or fail. The larger challenge is deciding which work humans should own, what skills that requires, and how you develop those skills in people who were hired for entirely different ones.
CIOs are not positioned to answer those questions. Their orientation is toward systems, not toward the human infrastructure that those systems are supposed to serve.
When AI strategy lives in the technology function, the questions that get answered are technical ones. The organizational and human questions get deferred, or delegated to HR as an afterthought once the implementation is already underway.
This is how you end up with an AI rollout that technically works and strategically fails. The tools function. The workforce does not know what it is supposed to be doing with them, or what its role will look like in three years, or whether to develop skills that the organization has not signaled it values.
This is not to say CHROs are currently rising to meet that responsibility. In fact, Swanagon would argue they aren't.
Most CHROs sadly are starting with, 'we want to automate things and we want to save money.' And it's good, but it's not fun," Swanagon said. "It's not interesting either. It's just good.
The M-shaped Gap
For years, talent strategy favored T-shaped workers — specialists with a wide enough view of adjacent domains to collaborate across them. The model made sense in an environment where specialization was stable and breadth was a coordination advantage.
AI is changing what breadth means. The emerging model is sometimes called M-shaped or comb-shaped: workers who have multiple areas of genuine depth, not just surface familiarity, connected by integrative judgment that lets them move fluidly between domains.
The distinction matters because developing comb-shaped workers requires a different kind of investment than developing T-shaped ones. You are not just widening someone's peripheral vision. You are building multiple competency centers in the same person and then cultivating the judgment to deploy them situationally.
Most organizations do not have a development program that does this. They have training catalogs and LinkedIn Learning licenses.
What Stewardship Requires
The word stewardship gets used loosely in leadership contexts. It usually means something like "we take care of our people" in the same vague register as "our people are our greatest asset." Neither phrase survives contact with a workforce reduction announcement.
Real stewardship in an AI transition means making decisions now that protect human value-creation capacity over a five- to ten-year horizon, even when those decisions are harder to justify in a quarterly review than another round of automation.
Recently, Adam DeRose from HR Brew and I sat down for an episode of Your Work Friends with Francesca Ranieri and Mel Plett. And Adam said something I spent the rest of the day thinking about.
“I understand the value of the quarterly business review, but honestly, I think we should just get rid of them.”
The reason being? The focus on quarterly performance has deteriorated leadership’s ability to think long term. Obsessions with quarterly numbers is pervasive, from board level all the way down to line managers, but the tactics that drive quarterly performance often undermine our ability to drive long term strategic thinking, especially around talent in the current moment.
This is a moment that just might require the funding the development of skills that do not yet have job titles. It means preserving the entry-level pipeline even when the math says you can eliminate it. It means treating the organizational question of what humans are for as a strategic one, not HR’s problem, and putting the people function in the room where those decisions get made.
Most organizations are not doing this. They’re optimizing. And optimization, applied to a structure that is changing beneath you, is a sophisticated way of running in the wrong direction.
The jobs being made more efficient today are not the jobs that will define organizational value in 2030. The work that will matter — judgment, navigation, ethical decision-making, the capacity to lead other humans through genuine uncertainty — is being developed and realized almost nowhere. It has no line item, no reporting metric, and certainly not enough executive sponsors.
That is the leadership failure. AI allows us to realize the goal of moving fast, but it’s worth asking if we’re breaking too many things along the way.
