- CollabWORK Hiring Insights
- Posts
- 54% of readers preferred the AI writing. Here's my complicated reaction as a former journalist.
54% of readers preferred the AI writing. Here's my complicated reaction as a former journalist.
What a viral NYT quiz is actually telling employer brand and TA leaders right now.

Hi everyone,
I’m sending this newsletter on the train to Philadelphia for IAMPHENOM. If you're there today, come find me; would love to connect in person!
This week, The New York Times published a quiz asking readers to compare passages written by humans or AI and vote on which they preferred. More than 86,000 people took the quiz and 54% liked the AI versions more.
I took it and picked the human passage 4 out of 5 times. As a former journalist who spent years working alongside Katie Couric — someone who built a career on the irreplaceable power of human storytelling — I'm the first person to say AI cannot replicate the nuance, the intellect, or the creative direction that humans bring and articulate.
But what this quiz is telling us is that people's appetite and willingness to trust AI-generated content is shifting. And that matters for hiring in ways I don't think the industry has fully worked through yet.
Employer brand is a human decision. Full stop.
Understanding a company's culture, speaking to C-suite leaders and translating business objectives into human capital strategy, building an EVP that actually means something — none of that can be handed to a model.
That work requires human insight and strategy and it always will.
But once that EVP exists, once a workforce shortage has been identified — whether that's radiologic technologists, where vacancy rates just hit an all-time high of 18.1%, or CDL drivers, where losing one can cost a fleet nearly $13,000 — the strategic work is done. What's left is communication and reactivation.
That's where I think most TA teams are leaving a lot on the table.
You've already done the hard part.
The candidates in your ATS and CRM already raised their hand: silver medalists, people who applied two years ago, former employees who left on good terms. They opted in, they gave you their email address, they joined your talent community…and most of them haven't heard anything useful from you since.
AI changes what's possible here. Not by replacing the strategy, but by executing on it at a scale and consistency that most teams can't sustain manually. You can take the employer brand your team built, the roles you've identified as critical, the story of why someone should come back or stay in your orbit, and actually deliver it consistently to the people who already know you.
A newsletter is the best place to start: consent is already there and the relationship already exists even if it's gone cold. And there's a compounding benefit worth naming: that content doesn't just live in someone's inbox. You can publish it to your career page or your company blog, and it becomes a fresh, indexable asset that AI models pick up when candidates ask "what's it like to work at X" or "who's hiring for Y right now." Re-engagement and employer AI visibility are the same work, done once.
I'll always believe humans break through the noise.
The NYT quiz didn't change my mind about journalism, storytelling, or the kind of work that moves people. It’s also why people need to show up consistently in the places candidates trust such as digital communities to form relationships that can’t scale through automated outbound.
But re-engagement isn't that work. Showing up consistently in someone's inbox with something relevant and useful — that's execution. And for execution, 54% of 86,000 people just told us the bar has moved.
If you want to think through what this looks like for your org, reply to this email. Looking forward to walking the floor at IAMPHENOM today and chatting more about this topic.
More next week.
Best,
Summer Delaney
CollabWORK Founder and CEO
Further Reading
Labor Market Impacts of AI —You've probably seen the viral graph floating around LinkedIn this week from Anthropic highlighting theoretical AI capacity versus observed usage. But the findings buried inside the research are what I keep coming back to: the workers most exposed to AI displacement are more likely to be older, female, more educated, and higher-paid. And while there's no systematic increase in unemployment for those workers yet, there is early evidence that hiring of younger workers in the most exposed occupations is already slowing.
Claude Has Its Own Substack — Anthropic is the latest brand launching a newsletter and unsurprisingly, Claude is writing it. My prediction for 2026: the employers building direct relationships with people who already want to hear from them will win on hiring efficiency.
If I Got Laid Off Tomorrow — I've been posting job seeker tips on TikTok and cross-posting the ones that land on LinkedIn. This is a recent video that got traction. The short version: one quality application a day beats 50 spray-and-pray ones, the referral is still the highest-converting path to a job, and the “hidden job market” continues to live in professional communities. Sharing it here because I know some of you might be forwarding this newsletter to people in a job search right now.