It’s time to have a different AI debate

Earlier this summer, I listened to Josh Bersin lay out a future where talent acquisition runs almost entirely on its own. Screening, scheduling, first-round interviews, assessment, offers. The technology has gone from assistant to agent to something that hears a business need and acts on it without much instruction. It was a genuinely impressive case, and it left me uneasy in a way I’ve been chewing on ever since.

Then a piece from Jason Averbook gave me the language for the unease. He wrote something that stopped me cold: we are spending enormous energy on how AI works, and almost none on how work should work. His frame is that every unnecessary task is like a bad hire. It consumes time, money, and attention without producing enough value to earn its keep.

Put those two ideas next to each other, and the automation question flips.

The debate we keep having is whether AI can replace the recruiter. That’s the wrong debate. The better question, the Averbook question, is which parts of recruiting were creating real value in the first place, and which were just motion we had gotten used to.

Because here’s the uncomfortable truth. A lot of what automation is about to erase probably should be erased. The seventh redundant screening call. The status update that updates nothing. The intake meeting that could have been three sentences. If a machine can do it faster and no one misses it, that work was never the point.

But some of it is the point. When a candidate is deciding between two offers and needs a straight answer about whether this team is actually as collaborative as the careers page claims, that’s not a task to optimize. That’s a relationship. When a hiring manager is talked out of a rigid requirement that would have screened out the best person, that’s judgment. Those moments don’t get faster with automation.

They get lost.

So the work worth protecting isn’t defined by what AI can’t yet do. It’s defined by what deserves a human even when a machine is available. That’s a values decision, not a technology decision, and it’s the one most teams are skipping.

My honest read for talent leaders heading into the fall: don’t start with the tool. Start with a list of everything your team does in a week, and put each item in one of two piles. Work that creates value, and work that just fills time. Automate the second pile without guilt. Then take the hours you got back and pour them into the first pile, the human parts, the ones that were always underfunded because everyone was too busy meeting about doing.

The companies that win the next few years won’t be the ones that automated the most. They’ll be the ones that were honest about what their work was for.