In partnership with
A friend told me over coffee last week that her company had given her a new title: “AI workflow lead.” Same job. What changed was that her manager now asks her every Friday what agents she had “delegated to” that week, and she has to send back a list.
She had no idea what to write the first two weeks. By the third week she was generating the list with ChatGPT-which felt, she said, like the kind of joke that wasn’t actually a joke.
I asked if she was doing the new job better than the old. She didn’t know. Then I looked at what was going on in the rest of the industry.
When AI Does the Work, Judgment Becomes the Job
In April last year Microsoft published its annual Work Trend Index, surveying 31,000 workers across 31 countries.
The report introduced a vocabulary built around what it called the Frontier Firm: companies running on hybrid teams of humans and AI agents, in fluid “Work Charts” instead of org charts, everyone an “agent boss”: someone who builds, delegates to, and supervises agents.
The most cited number: 82% of leaders said they expect to use AI agents to expand workforce capacity in the next 12 to 18 months.
Are you running your business on incomplete numbers?
Most small business owners have financials, but few have financial clarity. There's a real difference between books that are technically up to date and books that actually tell you what's going on in your business right now. When accounting is reactive — updated when there's time, reviewed at tax season — you lose visibility exactly when you need it most. You can't tell which clients are truly profitable. You can't spot a cash flow gap before it becomes a crisis. BELAY's outsourced accounting team changes that.
Two honest things about this source. Microsoft sells AI agent products. The vocabulary launched alongside Copilot feature releases.
And 82% is what leaders planned, not what they’ve done. The underlying trend is real. The branding is doing more work than the data is.
What is actually shifting, when you strip out the vocabulary: the most consequential skill in knowledge work is moving from “produce the work yourself” toward “specify the work, evaluate the output, decide what to keep.”
That is closer to editing than managing, closer to product taste than leadership. A junior analyst who could not previously tell a good chart from a misleading one cannot delegate chart-making to an agent and judge the result.
A senior analyst who could, can. AI delegation is not a new skill so much as a multiplier on an old one.
What’s next is almost here.
On July 16th at 1PM ET, beehiiv is going live with a look at the future of publishing, audience growth, and digital business.
What started as a newsletter platform has evolved into something much bigger: a place where creators and brands can grow, monetize, and own their audiences without stitching together half the internet to make it work.
The next chapter starts live at the Summer Release Event.
Join us to see what’s coming next.
The Microsoft Example
The example Microsoft features in its own report is a researcher there named Alex Farach, who runs three agents in parallel: one collects new research daily, one runs statistical analysis, one drafts briefs.
The setup lets Farach focus on the part of his job that requires judgment-which analyses are sound, what the briefs should actually argue. The example is a Microsoft employee in a Microsoft report; curated cases always look cleaner than the median.
But Farach’s setup works because he can evaluate every output. He delegates production and keeps judgment. If you cannot evaluate the output, you have not delegated work-you have generated workslop, just for yourself.
The smartest founders never pick their own 3PL.
Researching 3PLs yourself feels responsible. It's how you end up with a great sales pitch and a bad fit. There are over 2,800 of them, and the right one rarely markets the loudest. Here's why the founders who hand off the search get better answers, faster.
Try It This Week
One thing to do this week. Pick a recurring task where you can already tell a good output from a bad one-a type of analysis you’ve reviewed many times, an email genre you’ve drafted hundreds of.
Delegate one instance to an AI agent, then evaluate the output against your own standard. Keep what holds up. Cut what doesn’t.
The “agent boss” label is mostly marketing, but the underlying skill-knowing your own quality bar well enough to evaluate generated work against it-is the one that compounds.
—Prompt N Productive