The current state of AI in recruitment is not a great candidate experience
My beef with HR bots reading candidate slop
👋 Hey, it’s Diana! Welcome to this Operations Optimist newsletter. I write about building operations functions in startups, share my observations about venture capital and general musings about working in tech.
Everyone who has applied for a job in the last few years has lived this moment:
you hit submit on a meticulously crafted, tailored application. Weeks pass. And you wonder: did a human ever actually see that?
Increasingly, the answer is: no.
Our career aspirations are now filtered through an algorithm that decides if we’re worthy of a conversation — an entire ecosystem built on the illusion of efficiency.
Here’s why I’d rather read hundreds of human-crafted job applications than force you through AI-led recruitment process.
Everyone uses AI. So my bot is reading your slop.
ChatGPT rolled out more than three years ago. By now, everyone who has internet has probably used it. Every candidate is using AI to polish and tailor applications. Some lazy ones let ChatGPT generate the whole thing. Others (rightfully) use it to stress-test language and find right angles. I’m all for purposeful AI usage in a job hunt.
But this democratization of AI has completely wrecked the hiring pipeline. Companies are shutting down applications within 48 hours due to influx of applications.
Take my example:
I work at a 30-person tech company. We hire remotely. Our last non-technical job ad got around 700 applications. If I were to spend 8 minutes per application, work a 40-hour week, and take a few breaks during the day, it would take me 3 full weeks to screen. If you’ve ever worked in a startup, that’s basically forever.
So yes, the state of economy, remote work, and AI access have all flooded the hiring pipeline.
And in order to cope…
Meet the AI recruiter
Can you blame companies for using AI? Not really. Screening and scheduling takes forever. ATS systems have been around for long enough, and now that they’re AI-enhanced, they look like a perfect match.
The result?
Job applications have basically turned into keyword optimization. Candidates are no longer measured by capability or potential — they’re measured by how well they decode the hidden keyword structure of an ATS. It’s reached peak absurdity: LinkedIn lunatics are advising candidates to paste job descriptions or specific job related keywords into their resumes in tiny white font. The goal? Trick the machine into seeing a 100% match while the human eye sees nothing.
So, basically, now my bot is reading your slop. Maybe they are a perfect match after all.
AI interviews — the final stage of dehumanization
If ATS systems have been around, AI interviews are the new frontier. It’s basically a video interview with an AI agent that a hiring team watches asynchronously later. And again, I understand the need for scale.
But from a candidate’s point of view, it’s simply dehumanizing. Call me old-fashioned, but if a candidate took hours to craft an application, the recruiter can take 15 minutes to speak with a pre-qualified human.
How to use AI in recruitment for good
Don’t get me wrong, I think AI in HR can be amazing, just not at the parts we are now trying to automate.
I am excited for the world of AI-enriched candidate data with context like: Olivia is an avid user of your product and has just started her job application.
I also want LLMs to schedule meetings, summarize interviews, organize pipeline notes and remind me to send next-step emails, draft them and manage my calendar.
What I don’t want?
It making judgement calls on who makes it to the interview.
In the current economy and state of the job market, AI assessing and leading the hiring process is a purely frustrating candidate experience.
Let’s leave (for now) screening and interviews to humans. Honestly, I’d rather read 800 human-written applications than put a potential colleague through an AI-lead interview.
That’s it for today, thank you for reading!
PS. If you’re enjoying Operations Optimist, please help me make it more useful for you and fill in the reader survey.