Basic prompting is losing value. Workday's Global Workforce Report of October 5, 2026 found employer demand for simple prompting fell 25% after peaking in January 2026, while demand for building AI tools and automating workflows rose 51%. Dorie Clark's The Long Game explains the move: careers run in waves, and prompting was a reaping wave. Start the next learning wave by turning one repeated task into a workflow.

For two years the advice to anyone worried about AI was the same: learn to prompt. Job postings now say that advice has expired, and this week Workday put a date on it.

Workday's October 2026 Global Workforce Report, released on October 5, 2026, analysed the skills named in job requisitions at roughly 550 enterprise employers using Workday Recruiting. Demand for basic AI skills such as simple prompting rose through late 2025, peaked in January 2026, and then fell 25%. Over the same period, demand for hands-on skills (building AI tools, automating workflows, AI engineering) climbed 51% between September 2025 and July 2026. The report also surveyed 6,001 employees and business leaders: 40% of leaders expect AI to get more out of the people they already have, and only 28% expect it to cut headcount.

Two other numbers in the report matter more to a founder or executive than the headline. Mentions of management and leadership skills in job postings fell 7%, and mentions of training skills fell 13%, between September 2025 and July 2026. And while 79% of workers say they know which skills they need, only 66% say their employer helps them build those skills. Companies are asking for builders while cutting back on the people who teach. If you run a team, that gap is yours to close. If you work alone, nobody is going to close it for you.

Workday's August AI@Work Pulse survey of 5,944 workers explains why prompting stopped being scarce. 57% say they cannot rely on AI output without checking it closely, and 33% say at least half of the time AI saved them is later eaten by reworking low-quality output. A prompt gets you a draft. The value has moved to what happens around the draft: the workflow that runs it again tomorrow, the check that catches the error, the decision about whether the task should be automated at all. That is what "building" means in these postings, and for most operators it does not mean writing code. It means turning a repeated piece of your work into a system that runs without you re-explaining it every time.

Dorie Clark's The Long Game is the right lens here. She argues that careers move in waves rather than straight lines: Learning, Creating, Connecting, Reaping, and then back to Learning. Her warning is about the fourth wave. People who find something that pays keep reaping it and stop investing in what comes next. Prompting was a reaping skill for about eighteen months, and the January 2026 peak in Workday's data looks like the crest of that wave. Clark has a second tool for the cost of starting over: decide what to be bad at. You cannot learn to build workflows while also staying the fastest prompt-writer in the room, and you do not need to.

This is the protocol I would use to move from the prompting wave to the building wave without stopping the work that pays. Call it the repeat-to-workflow shift:

  1. Audit the repeats for one week. List every AI conversation in which you typed instructions you have typed before. Anything that appears three or more times is a candidate.
  2. Build one, small, in your 20% time. Clark's 20% rule, applied narrowly: take the most frequent repeat and turn it into something that runs from a saved instruction set, a scheduled automation, or an agent with fixed inputs. One finished workflow beats five half-built ones.
  3. Write the check before you trust it. Given that 57% of workers cannot rely on output unchecked, write down what a wrong result looks like and how you will catch it before the workflow runs unattended.
  4. Decide what to be bad at. Stop hand-polishing one output and let the workflow's slightly worse but repeatable version stand.

The limits are real. Skill mentions in job postings measure what recruiters write, not what work pays, and a 25% fall from a January peak still leaves prompting as a baseline expectation rather than a worthless one. The employer data comes from Workday's own customers, which skews toward large companies, and Workday sells AI agents for HR and finance, so it has a stake in the story. "Build with AI" also covers everything from a saved instruction set to AI engineering, and the 51% rise does not say which end employers mean. Treat the report as a direction, not a curriculum. The direction is clear enough: the skill that is getting cheaper is asking well, and the skill that is getting more valuable is making the asking unnecessary.

Sources: Workday Global Workforce Report: AI Is Rewriting Jobs More Than It's Cutting Them, Workday Newsroom, October 5, 2026; New Global Workforce Report: 5 Takeaways Leaders Can Act on Now, Workday Blog, October 2026; AI Adoption Is Moving Faster Than Employee Trust (AI@Work Pulse, August 2026 survey), Workday, September 2026; Dorie Clark, The Long Game: How to Be a Long-Term Thinker in a Short-Term World (Harvard Business Review Press, September 21, 2021), chapters on Career Waves, 20% time and deciding what to be bad at.

Related: How to Find Your Passion · Best Self-Improvement Books · How to Make Better Decisions · What University Will Not Teach You