The Reader Who Isn't a Person

Carl Spitzweg, "The Bookworm" (1850). We picture a careful human reader. For many first jobs, that reader is now software. Public domain, via Wikimedia Commons.

That sounds strange, so let me explain. Most companies now use software to sort résumés before any human ever sees them. More than nine in ten employers rely on some kind of automated system to filter or rank the people who apply, according to a 2021 study led by Joseph Fuller, a professor at Harvard Business School.^[1] The software scans your résumé, gives it a score, and decides whether you move on.

This first reader is not looking for warmth or a clever sentence. It looks for keywords, clean formatting, and skills that match the job posting. The friendly cover letter that took hours to write is, to the software, mostly invisible.

And when a human finally does look, they do not look for long. On average, a recruiter spends about 11 seconds on a résumé that makes it past the filter.^[2] Eleven seconds is not really reading, just a quick glance to confirm what the software already picked.

So my nephews were never writing for a person. They were writing for a machine, and no one had taught them how that machine reads.

The Skill No One Teaches

William Powell Frith, "The Derby Day" (1856–1858). A vast crowd of individuals, each one now sorted by the same few systems. Public domain, via Wikimedia Commons.

I believe this is the most important skill for the next ten years, and it is not coding. I call it algorithmic literacy, which is understanding what a machine is looking for, and learning to speak its language without losing who you are.

In 2026, Rishi Bommasani, a researcher at Stanford, led a study of 3.4 million real job seekers, the largest of its kind ever done.^[3] They discovered that many different companies use the same few software programs to sort applicants. Because of that, one person can be turned down again and again, by separate employers, for the very same reason. A "no" from one program can quietly become a "no" in dozens of places at once.

Apply to ten jobs, and there is a real chance the same kind of software screens you out of all ten before a single person ever sees your name. That is not ten unlucky coin flips. It is closer to one decision, repeated.

The new skill of the decade is not coding. It is learning to read the machines that read you first.

The Machine Is Not Always Fair

There is one more thing worth understanding. These programs learn from old data, and old data can carry old bias. In a 2024 study, Kyra Wilson, a researcher at the University of Washington, and Aylin Caliskan, a professor there, asked an AI to rank thousands of résumés.^[4] It favored names linked to white applicants far more often than names linked to Black applicants. The machine is not neutral, and knowing that is part of being ready for it.

Bigger Than a Résumé

The résumé filter is only the beginning. The same kind of software is starting to decide who gets a loan, an apartment, or a place in a program, and the résumé is simply the first machine most of us have to get past.

Carel Fabritius, "The Goldfinch" (1654). Miners once carried a caged bird underground to warn them when the air went bad. Public domain, via Wikimedia Commons.

Think of the canary that coal miners once carried underground. The little bird warned them when the air turned dangerous, long before a person could feel it. The résumé filter is our canary, an early warning about a world where machines read first and people read second.

What You Can Actually Do

So what can you actually do about it? Here is the short version I gave my nephews.

Keep it clean and simple. Use a plain résumé with one column and normal headings like Experience, Education, and Skills. Skip the fancy graphics and text boxes, because the software can scramble them and lose your information.

Match the job post. Read the exact words in the listing, then use those same words yourself. If the post asks for "inventory," write "inventory," not "kept the back room tidy." Ten résumés that match the posting will beat a hundred that don't.

Find a real person. The most reliable way past the software is to not face it alone. A referral, a short message to a manager, or a five-minute chat at the counter can matter more than fifty online forms.

Ask what the machine wants. For any system you deal with, ask a single question: what is this trying to measure, and what is it using to stand in for me? That habit will protect you long after this job hunt is over.

Stay the Author

I am going back to my nephews with all of this. Together, we are going to rebuild their résumés in plain text, pick a dozen jobs instead of a hundred, and put real effort into reaching actual people. None of it promises a job, but all of it improves the odds.

The work never changed, and my nephews never changed. What changed is that they now understand who, and what, is reading. That understanding is the skill of the next ten years, and almost no one is teaching it.

The first machine to read you should not get to write your ending. Learning its language is how you stay the author of your own story.

Disclosure: This article is published by Sage.Education, a product of Startr LLC. The views expressed are those of the editorial board.