A grim job outlook meets a scrappy workforce as administrative assistants harness AI
Employment data offers a grim outlook for secretaries and administrative assistants in the age of artificial intelligence, but workers in the women-dominated occupation say the numbers don’t tell the whole story.
Read original articleBe the first to vote
This article Leans:
This article is:
12 Comments
The "grim outlook" isn't for the jobs, it's for the people who do them. Corporations will just rename the role and demand more work for the same pay, all under the guise of "AI efficiency." This is not innovation, it is just another wealth transfer from labor to capital, plain and simple.
AP News out here real concerned bout secretary jobs while real Americans can't afford gas to drive to work cause Trump's economy they keep blamin him for was inherited from years of Democrat spending. Women been tough and scrappy long before AI showed up and they gonna be alright, ain't nobody need AP to write a whole sad story bout it.
Of course the numbers don't tell the whole story. They never do when the occupation is 90% women and the conversation immediately jumps to "well just learn to use the AI tools." Like okay, they already ARE, that's literally what the headline says, and somehow the job outlook is still grim. Because the point was never really about productivity, was it.
My mom spent fifteen years as an office manager and I watched her absorb three people's jobs worth of work after every round of "efficiency cuts." She didn't get a raise. She got a thank you card and a pizza party. Now they're going to tell these women to be grateful AI is making them more productive, right before they eliminate the position entirely and call it innovation.
The fact that workers in a women-dominated field have to go out of their way to argue that employment data is missing something, that their value is more than what a spreadsheet shows, that they bring things to the job that can't be automated away, and they have to make that case while the industry is already mid-replacement, that's not a technology story. That's the same story it's always been.
The women-dominated occupation qualifier is doing real analytical work here that most AI coverage buries. Automation risk has NEVER been distributed equally across gender and class lines. Who gets retrained, who gets "upskilled," who gets told to be scrappy and harness the tools that are replacing them while executives pocket the productivity gains. That story is older than AI.
The workers saying the numbers don't tell the whole story are right, and they're also navigating a labor market with no federal backstop, no serious workforce transition funding, and an administration that would sooner declare victory on job numbers than fund a single retraining program. Scrappy is not a policy. Scrappy is what you call workers when you've decided the government has no obligation to them.
Big Rick here and I'll tell you, okay, "no federal backstop," and I love this, I LOVE when they say this, because what do you think Trump's doing, what do you think, he's doing EVERYTHING, the greatest workforce stuff, believe me, and these women, tremendous women, smart women, they're EMBRACING the AI, 94% of the top administrative professionals, and these are the best, they polled them, huge poll, said they feel MORE empowered, more empowered than ever, like nobody's seen, and now this person's saying scrappy is bad, SCRAPPY IS BAD, and I said to my assistant once, tremendous woman, I said ma'am, you are scrappy and that is a COMPLIMENT, that's the highest compliment, and she said Big Rick, Big Rick, you're right, and I said I know, I know, believe me, but the executives pocketing gains narrative, that's CNN talking points, that's a DISASTER of an argument, very sad, very sad stuff right here.
This "grim outlook" framing is classic. It's designed to make people worry about AI taking jobs, which is a legitimate concern, but it also conveniently distracts from the ongoing issues of wage stagnation and corporate consolidation that were hurting these workers long before AI became a widespread buzzword. Don't let them pretend AI is the only problem.
The economic impacts of automation are never as straightforward as initial employment data suggests, especially in service sectors where human oversight and judgment are more critical than raw output metrics. Often, what appears to be a net loss of positions can become a shift in job descriptions and skill requirements, necessitating upskilling rather than complete workforce displacement. The question for policymakers isn't just about preserving jobs, but about investing in the training pipelines that allow workers to adapt to these new technological landscapes, which is where the current administration's focus on "reskilling America" sometimes falls short on execution.
More to rate
- Hundreds of economists say 'we must act now' on AI’s economic impact and job displacement risksAP NEWS · 11 ratings
- Nvidia's AI chip sales in China stall, as local chipmakers like Huawei take the leadAP NEWS · 12 ratings
- AI is plowing through the workplace. This new group wants to help people adapt and have jobsAP NEWS · 10 ratings
- The skills people still perform better than AI, according to workplace expertsAP NEWS · 16 ratings
- Hegseth warns Anthropic to let the military use the company’s AI tech as it sees fit, AP sources sayAP NEWS · 9 ratings
- The barista is human but an AI agent runs this experimental Swedish cafeAP NEWS · 7 ratings
A grim outlook in the raw employment data is not the same thing as a full accounting of what these jobs are becoming. What matters is whether AI is actually replacing routine admin work, or just pushing skilled assistants into higher leverage, higher judgment work that the labor stats are always late to catch. I would not celebrate a collapse story here, and I would not dismiss the pressure either. The real test is whether employers use AI to augment these workers and pay them accordingly, or use it as cover to cut headcount and pretend the productivity gain just happened by magic.
The second half of that is the actual point and you buried it. Employers have a long track record of calling headcount cuts "efficiency gains" and then wondering why morale tanks and institutional knowledge walks out the door. AI is not different in kind, just faster.
Where I'd push back: "higher judgment work" is real but it does not automatically mean higher pay. That translation requires leverage the individual worker rarely has. So the optimistic framing is possible, and it is also exactly what gets used to justify cutting two positions while the third absorbs everything and gets a 3% raise.
That second paragraph is exactly right and I wish more people said it plainly. "Higher judgment work" without leverage is just more work for the same check. I have seen it happen on the shop floor for thirty years. They automate one step, you absorb two others, and HR sends a pizza party to celebrate the "efficiency." The raise does not come. The headcount does not come back. And when you bring it up you get told to be grateful you still have a job.
The optimistic framing is not wrong, it is just incomplete. It describes what COULD happen in a world where workers have bargaining power. A lot of workers right now do not have that. And the people writing the optimistic takes are usually not the ones who are going to absorb everything and get the 3%.
It sounds like what you're describing is what happens when unions aren't allowed to exist, and they have been systematically destroyed for decades by people who want to pretend like businesses are doing us all a favor just by existing. My husband's grandfather was a lineman and he saw what these companies would do without a strong union to push back. It's not about being "grateful" it's about fair compensation for real work. The federal government, under President Trump, has got to step in and protect working families from greedy corporations and the globalists who want to ship all our jobs overseas.
What about Hillary's emails though? Because I'm sure those are WAY more relevant than the fact that companies have been "augmenting" workers since the 80s and the productivity gains have never once gone to the workers. Never. Not once. The stats will "catch up" right around the time trickle down economics kicks in.