How Recruiters Can Beat AI On Evaluating Candidates

Artificial intelligence has already gotten disturbingly good at reading people, from consoling the upset to diagnosing patients with more empathy than trained professionals. That should worry anyone whose job depends on judging candidates. But most hiring managers are still such weak interviewers that AI wins by default. A skilled interviewer, using the right questions, can still beat it. The material draws on research tracking tens of thousands of new hires to figure out why so many failed within their first year and a half. The answer was rarely a skills gap. It was attitude. Skills are easy to test: give a coder a bug to find, an analyst a spreadsheet to build. Attitude is harder, and it's the piece most interview processes never actually measure. Key points covered include:

• Why organizations should stop chasing the perfect hire and instead focus on never hiring a "talented terror," the candidate with strong skills and a toxic attitude who passes every test but causes outsized damage once inside the building.
• Why classic interview questions like "tell me about yourself" or "what's your biggest weakness" are worthless, since every candidate walks in with a rehearsed answer.
• Why hypothetical questions ("what would you do if...") only reveal what someone thinks they'd do, not what they'd actually do under pressure.
• Why quirky personality-test questions, including one real example built around what animal a candidate would be, have no scientific basis and often reward the wrong traits entirely.
• The habit of tacking phrases like "and how did you overcome that" or "and did so successfully" onto behavioral questions, which quietly hands candidates the answer key by signaling exactly what kind of story to tell.
• A simple fix: ask about the problem itself, not how it was solved. Problem solvers reveal themselves unprompted; problem bringers expose themselves just as fast.
• Specific rewrites of common questions, stripped of the words that give away the desired outcome, so interviewers hear what candidates actually did, or didn't do, when things went wrong.

The larger point is that eliminating the bottom half of failed hires, particularly the talented terrors, does more for a team's performance than any effort spent trying to find a flawless candidate.