Artificial Idea | AI careers · practical prompts · no hype Monday, October 6, 2025 · Issue #19 · Jobs

The promotion problem

The professionals who are not being promoted are not underperforming. They are performing exactly as asked. That is the problem.

There is a pattern visible in organisations across every sector that is producing a specific and largely unacknowledged form of professional frustration. It goes like this. A team adopts AI tools and becomes measurably more productive. The team delivers more output with the same headcount. Senior leadership notices the productivity gain and draws the obvious conclusion: the team can deliver even more, or the same amount with fewer people. Headcount is reduced or held flat as scope expands. The remaining team members absorb the additional work, enabled by the same AI tools that created the efficiency gain in the first place.

The professionals inside that pattern are working harder than they were twelve months ago, delivering more than they were twelve months ago, and in many cases earning exactly what they were earning twelve months ago. They are doing more with less, as the management consultants say. What the management consultants rarely add is that doing more with less, when the less refers to resources and the more refers to your own labour, is a value transfer from employee to organisation that does not automatically come with compensation on the other side.

This dynamic has a name in the organisational psychology literature. Job expansion without job enrichment. It is distinct from job enrichment, which involves adding responsibilities that develop capability and increase professional value. Job expansion adds volume without adding value, complexity without adding development, and output requirements without adding the resources or recognition that make the expansion worthwhile.

AI is accelerating job expansion in a specific and measurable way, and the professionals most at risk of being trapped in it are those who are genuinely good at their current roles, genuinely committed to their organisations, and genuinely capable of absorbing the additional work AI enables without visibly breaking. Their competence is being used against them.

The research on what is actually happening

A 2025 survey of 3,400 knowledge workers across the United States, United Kingdom, Germany, and India conducted by Mercer found that 67% reported their workload had increased over the previous twelve months. Of those, 71% attributed the increase at least partially to AI-enabled productivity expectations from their organisation. Only 23% reported a corresponding increase in compensation. Only 18% reported a corresponding increase in seniority or scope of responsibility.

The gap between productivity contribution and professional advancement in that data is not a rounding error. It is a structural dynamic that organisations are not, for the most part, managing intentionally. They are benefiting from it without having designed it, and in many cases without having fully acknowledged it internally.

A separate 2025 analysis by Deloitte's Human Capital practice found that employee engagement scores at organisations with high AI adoption rates declined by an average of 14 percentage points in the eighteen months following tool deployment, with the sharpest declines concentrated among mid-level professionals in roles that had seen the greatest productivity gains. The interpretation Deloitte offered was direct: organisations were capturing the productivity dividend of AI adoption without sharing it in ways that employees found meaningful, and the employees most affected were those whose productivity had increased most.

The professionals experiencing this dynamic are not wrong to be frustrated. The frustration is a rational response to a real asymmetry. The question is what to do about it productively, because frustration that stays internal produces nothing except attrition risk, and attrition in the current labour market is a more complicated calculation than it was three years ago.

The distinction between job expansion and job enrichment

Understanding this distinction clearly is the precondition for doing anything useful about the dynamic described above.

Job expansion is what happens when AI increases the volume of work you can produce and your organisation responds by expecting you to produce more volume. You are doing the same things, faster, at greater scale. Your skills are not developing. Your strategic value to the organisation is not increasing. Your compensation case is not strengthening. You are running faster on the same treadmill.

Job enrichment is what happens when AI handles the volume-intensive, routine components of your role and you redirect the recovered time toward work that develops your capability, increases your strategic visibility, and builds the kind of professional track record that supports advancement. You are doing different things with the same hours. Your skills are developing. Your strategic value is increasing. Your compensation case is strengthening.

The difference between these two outcomes is not primarily determined by what your organisation does. It is primarily determined by what you do with the time AI returns to you. Organisations that are expanding jobs rather than enriching them are not, in most cases, explicitly instructing their employees to absorb more volume. They are creating conditions in which absorbing more volume is the path of least resistance, and most employees, being conscientious and committed, take the path of least resistance without examining where it leads.

The professionals who avoid job expansion trap take a different posture. They treat the time AI returns to them as a resource to be allocated strategically, not a buffer to be filled with more of the same work. They make that allocation visible to their managers by explicitly proposing how they intend to use the recovered capacity, rather than quietly absorbing additional volume and hoping the contribution is noticed.

Hoping contribution is noticed is a career strategy with a poor track record at every level of organisational seniority.

How to push back without pushing back

The framing of pushing back is worth examining, because it implies a confrontational dynamic that is rarely the most effective approach and rarely accurate to the actual situation. Most managers who are expanding their team's jobs are not doing so maliciously. They are doing so because the path of least resistance for them is to increase output expectations when productivity increases, and nobody in their team has given them a reason to think about it differently.

Giving them a reason to think about it differently is the intervention. It does not require a confrontation. It requires a conversation, framed in the language that organisations actually respond to, which is the language of strategic value rather than workload fairness.

The conversation has a specific structure that is worth understanding before having it.

It begins with an acknowledgment of what AI has enabled: the productivity gains are real, the increased output is real, and you are not disputing either. It proceeds with a proposal: given the recovered capacity, you would like to redirect a defined portion of it toward a specific higher-value activity, project, or development area, with a clear articulation of why that activity is more strategically valuable to the organisation than additional volume of the same work. It ends with a request for explicit agreement on how the recovered capacity is allocated, so that the expectation is managed rather than left to accumulate without discussion.

This conversation does two things simultaneously. It prevents the job expansion trap by creating an explicit agreement about what you are doing with recovered time. And it signals to your manager that you are thinking about your role in strategic terms rather than operational ones, which is the signal that most reliably precedes advancement conversations.

A 2024 study published in the Journal of Organizational Behavior found that employees who proactively proposed reallocation of their time toward higher-value activities following productivity improvements were 2.8 times more likely to receive a promotion within eighteen months than those who absorbed additional volume without comment, even when the volume absorbers produced objectively more output. Output volume is not the variable organisations promote on. Strategic initiative is.

The specific ask that works

The conversation described above is most effective when the higher-value activity you propose is specific, connected to an organisational priority that your manager cares about, and framed as an experiment with a defined timeframe rather than a permanent reallocation.

Vague proposals to "focus more on strategic work" produce vague responses. Specific proposals to spend eight hours per week over the next quarter leading a particular initiative, developing a particular capability, or producing a particular output that your manager values produce specific responses, including sometimes a no, which is also useful information.

The no tells you either that your manager does not value the proposed activity, in which case you have learned something important about the alignment between your development priorities and your organisation's current direction, or that the timing is wrong, in which case you have opened a conversation that will be easier to reopen next quarter than it would have been to start from scratch.

Either outcome is more useful than the alternative, which is absorbing additional volume indefinitely while wondering why advancement is not coming.

The action

Identify the single highest-value activity you are currently not doing because the time required for it is being absorbed by volume work that AI could handle or has already enabled you to handle faster. It should be specific: a project, a capability, a relationship, a piece of analysis, a visible contribution to an organisational priority.

Then schedule the conversation with your manager in which you propose reallocating a defined portion of your recovered capacity toward it. Not a complaint about workload. A proposal about strategic allocation. The distinction in framing produces a different conversation, a different response, and over time a different trajectory.

The professionals who are advancing through the current AI transition are not the ones doing the most work. They are the ones doing the most strategically visible work. Those are different things, and the gap between them is where careers either develop or stall.

Thursday we are giving you the prompt framework that turns exactly this kind of strategic conversation into something you can prepare for rigorously rather than approach with good intentions and insufficient preparation. The prompts cover how to frame the proposal, how to anticipate your manager's likely objections, and how to follow up in a way that keeps the conversation moving rather than letting it die in someone's inbox.

Preparation is not a substitute for the conversation. It is what makes the conversation worth having.

— The Artificial Idea team