Interviewing after AI in Jan 2024

This post is about empathy and adapting as a hiring manager to our post-AI world.

Hi. Is winter over yet? Can I come out of hibernation? Can I stop being depressed? No? Okay. Well, I’m gonna do this anyway.

This post is about empathy and adapting as a hiring manager to our post-AI world.

I’m running interview pipelines for the first time in about a year and the experience has changed significantly since then. I am seeing many parts of the application are being run through Large Language Models. I first noticed this when an applicant sent a cover letter with placeholders left in: it literally said “in my time at [Previous Company]”. That caught my eye and I spotted their cover letter insisting they knew tech from our job advertisement that their CV didn’t mention! No Bueno.

Fine, cover letters were arguably always a waste: a few paragraphs of echoed buzzwords, a game to demonstrate you read the job description. Now candidates put the ad into a prompt with “please write me a cover letter”. LLMs have made cover letters obsolete. Finally!

But candidates are also writing their CVs with LLMs. I’m starting to notice a trend in language and tone, certain similarities in style and word choice cropping up. I want to be clear: Using AI is fine! Not everyone knows how best to express their history, their work or their successes in writing. We’ve always got help from a friend, mentor or peer—I’ve helped plenty of folks with ideas, words and phrases. Now people without experienced friends have access to support. Brilliant!

The part that’s not so brilliant is that current LLMs embellish. They know what people want to read in CVs. If you trust the software and you’re not good at writing, you might not edit it either (placeholders aside.) This is where my issue arises: if I’m looking for people who have all three of “developed, built and deployed using tech X”, your CV says “I developed, built, and deployed using tech X” and it comes out in the interview that you didn’t actually do all of those … I’m not going to be very happy because we’re wasting everyone’s time. This is, essentially, a lie. A CV is an information dense document where the individual words matter, and a stated fact is inaccurate.

In the past I heard from folks who worried that structured and open interviews wouldn’t catch liars out. I’ve always said that someone going out of their way to misrepresent themselves in the interview is going to be extremely rare, and it’s generally not worth your time to try and catch those people out (I’ll write more in the future about what we need to do instead: clear and consequential probation’s.)

What this new paradigm is doing is making inadvertent liars of people. With good intention they copy and paste, and suddenly they’re lying. There’s no malice here, rather there’s an innocent and authentic wish to communicate. I have compassion for that. But it means that as hiring managers and interviewers, we now need to catch these inadvertent liars as early as possible.

Some of the work for us here is philosophical. To me lying on your CV is a deal-breaker—I don’t want to hire you if you’re willing to misrepresent the truth. But this is accidental, almost innocent. I can’t paint these candidates with the same brush as the conscious manipulator of yesteryear. I need to change my approach.

So what do we do about this? How do we not only hire candidates who are able to fulfil the obligations of the role, but also not waste our time interviewing? It’s going to need more empathy and openness, not less. Welcome to 2024, gang. It’s going to be exciting!