AI Doesn't Give You Better Candidates

One day, I recommended a candidate for a position at work to my recruiter. Before I could even get back to my desk, I received a curt reply: "Sorry, this won't work. We aren't hiring people from USC anymore."

That became the running joke in the office: "You got lucky; we hired you before we realized how bad you were."  

The reasoning behind this decision wasn’t exactly nuanced. The company, known for its high turnover rate, decided USC graduates weren’t worth hiring because they didn’t perform well or stick around. But digging deeper, the issue wasn’t with the candidates. It was with us.  

This was a company that offered low pay and demanded high-quality work under intense conditions. Naturally, recruiters gravitated toward fresh graduates with no work experience because they were "affordable." The result? We churned through USC grads who couldn’t meet the unreasonable expectations we set for them.  

The irony was that every now and then, we stumbled upon exceptional employees from this pool. These rare successes gave the illusion that USC was a solid candidate source. But it was akin to finding water in the desert. Possible, but not the most effective strategy.  

The reality was simple: the system was broken, and blaming USC graduates was a way to avoid fixing it. Instead of adjusting our pay, expectations, or recruiting strategies, we wrote off an entire group of people as lazy or unqualified.  

This is where AI often enters the story, pitched as the magical solution to hiring problems. But here’s the thing: AI doesn’t create better candidates. It just optimizes the process of selecting from the same pool of applicants. If the pool itself is flawed, whether it’s due to systemic biases, unrealistic expectations, or poor outreach, AI will only reinforce those flaws.  

Take the story of a manager whose CV was rejected by his own company's AI-driven applicant tracking system. He discovered the system automatically filtered out every applicant, including himself, because it was prioritizing outdated AngularJS skills instead of the modern Angular expertise actually required for the role. The result? Qualified candidates were rejected en masse, and the HR team was eventually fired for relying too heavily on the flawed system without ever reviewing applications manually.  

The same principle applies to dating apps. No matter how sophisticated the algorithm, the app can only match you with people who’ve signed up. Paying for premium features or enhanced matching doesn’t magically add better people to the pool. It just filters the same group in different ways.  

For hiring, the lesson is clear: you have to do the work. AI can help streamline and automate parts of the process, but it cannot compensate for bad inputs. If you want better results, you need to address the real problems. Widening your reach to attract more diverse and qualified candidates, offering competitive pay, and creating a work environment people want to stay in.  

In the end, technology is a tool, not a solution. AI doesn’t fix broken hiring practices. It just makes the flaws harder to ignore.

---