Instead of reducing work, AI will increase expectations.
Courtesy: Gautam Shiv, with Tejwansh S. Anand
There is no shortage of opinion on what AI will do to the workforce. Economists publish displacement curves. Vendors release productivity benchmarks. Consultancies forecast which jobs will disappear, and by when. I have read a lot of this commentary over the past year, and at some point I started noticing what was missing from almost all of it: the voice of the people who are actually about to live through the transition.
This spring, I had a chance to go looking for that voice directly.
As a Teaching Assistant for a graduate course called Harnessing AI for Business at the University of Maryland’s Robert H. Smith School of Business, I watched 124 students complete an assignment that asked them to analyze how AI would change their own chosen career path. Business analysts, consultants, data scientists, product managers, cybersecurity professionals, and more, each one researching their own future.
When the submissions came in, my professor was impressed enough that he asked me to read every one of them. So I did. All 124. And what I found was not what I expected.
There was no panic. No hype. Just a remarkably consistent, clear-eyed read on what is actually changing in professional work and what it will take to stay ahead of it.
A Pattern That Held Across Every Career
“98% of students reached the same core conclusion, independently of each other: AI will transform their role. It will not eliminate it.”
That number is worth sitting with for a second. These were students analyzing completely different professions, citing different sources, writing without seeing each other’s work. A cybersecurity student and a future consultant landed in the same place. So did a data analyst and someone headed into HR.
Business analysts, data roles, and consultants made up the largest share of the 124 analyses.
But the conclusion that actually changed how I think about this topic came from a single line, written by a student analyzing product management.
“AI will not eliminate my career path, but it will significantly change what the role looks like. Instead of reducing work, AI will increase expectations."
The phrase “increase expectations” turned out to describe something 75% of students identified on their own, in their own words. I started calling it the speed paradox.
The Speed Paradox
Here is the assumption most of us carry into conversations about AI and productivity: if a tool makes a task ten times faster, you get time back.
That is not what these 124 students found.
What they described instead is that organizations recalibrate expectations just as fast as the tools improve. If AI makes something ten times faster, the output expected of you rises by roughly the same amount. The bar does not stay still while the effort to clear it shrinks. It moves with the tool.
This hits hardest at the entry level. The tasks that once taught junior professionals the ropes, the first draft reports, the routine research, and the repetitive analysis are exactly the tasks AI now handles. The on-ramp into a career is getting shorter right as the expectations for new hires climb higher.
If you are early in your career, this is the single most useful thing in this entire piece.
The Skills Trade
90% of students described the same shift in what makes someone valuable at work, no matter which field they were analyzing.
Declining in value:
Manual data entry and cleanup
Routine report writing
Basic, repetitive coding
Following a fixed process without adapting it
Rising in value:
Strategic framing and judgment
AI literacy and the ability to work with these tools well
Ethical reasoning and bias detection
Critically evaluating AI-generated output
Domain expertise that cannot be reduced to a prompt
This tracks closely with the World Economic Forum’s 2025 Future of Jobs report, which named critical thinking and creativity as the fastest-growing skills employers need. The difference is that our students reached the same conclusion from inside their own career planning, without ever reading that report.
More Data, Not Less Need for Analysts
Among students analyzing data-heavy careers, a different and slightly counterintuitive pattern showed up. The common assumption is that AI, by automating routine analysis, will shrink the need for human analysts.
68% of these students argued the opposite.
As AI makes basic analysis faster, organizations ask more questions and generate more data than before. The bar for what counts as a meaningful analytical contribution keeps rising. Demand shifts toward the people who can frame the right question, catch an AI-generated finding that is subtly wrong, and turn complexity into a decision someone can actually act on.
One student summed it up in a line I have not been able to stop thinking about:
How It Plays Out, Career by Career
Business analysts described their value as relational, not mechanical. AI can draft requirements documentation. It cannot negotiate alignment between two stakeholders who disagree or read the politics of how a decision actually gets made. Several expect the role to become more strategic as the documentation work shrinks.
Consultants were the most direct about what AI takes over: research, slide generation, first draft frameworks. What stays, in their view, is client trust and the judgment to adapt a recommendation to a specific room. One student put it perfectly: AI can build the deck. It can’t sell the recommendation.
Product managers were the most optimistic group I read. Several pointed to faster prototyping, the ability to go from idea to working demo in days instead of weeks. They also flagged something genuinely new: designing products where humans and AI collaborate well, a skill that did not exist in PM job descriptions two years ago.
What I Took Away From This
If you are an employer, the entry-level compression problem deserves real attention. The tasks that once trained junior employees are disappearing into automation just as expectations for new hires rise. That gap does not close on its own.
If you are an educator, like the professor I work for, the question worth asking is whether learning outcomes are keeping pace with the workforce students are actually entering. These 124 students had already started adapting on their own, using tools like ChatGPT and Claude that were never assigned in any course.
And if you are early in your career or about to be, the strategy implied by every one of these 124 essays is the same: build AI literacy on purpose, go deeper on the expertise that makes good use of these tools possible, and treat evaluating AI output critically as a core skill, not an afterthought.
A Quick Honesty Check
This is 124 students from one graduate program at one university. It is a meaningful dataset, not a representative one. A broader study, across multiple schools and including professionals further into their careers, would tell us whether this pattern holds beyond one classroom. That is exactly what I want to pursue next.
What this research does offer right now is something the public conversation about AI and work has mostly lacked: a direct, first-person account from people who are about to live the answer instead of debating it from a distance.
The Line I Keep Coming Back To
One student, writing about her future in product management, ended her essay with this:
Based on everything I read across these 124 essays, the next generation of business professionals is not waiting around to find out what happens next.
They are already writing it.
Gautam Shiv is a teaching assistant for BUDT751: Harnessing AI for Business at the University of Maryland’s Robert H. Smith School of Business, where he completed his MS in Information Systems in December 2025. Tejwansh S. Anand is a professor at the Robert H. Smith School of Business and the instructor of BUDT751.
Sources referenced: World Economic Forum, Future of Jobs Report 2025; HEPI, Student Academic Experience Survey 2025; OECD, Digital Education Outlook 2025.