How I Sniff Out Your BS

Up until mid-2024, I was convinced my hiring track record proved I was incompetent.

I told myself I should never be trusted to hire again.

For a long time, I couldn’t figure out what was going wrong…smart-seeming people in interviews who inevitably crumbled once they were in the role.

Then I pinpointed it: remote work + AI hadn’t just changed hiring. They had amplified deception in the hiring process.

Cheating, outsourcing thinking, and lack of ownership in the hiring pool were always problems. But after COVID and AI’s rise, it swung to unprecedented extremes.

The root cause of my every hiring failure: Outsourced thinking. And my inability to catch it.

Outsourced thinking means relying on external sources or shortcuts, whether it's AI-generated answers, half-baked input from others, or simply skimming through content without actually processing it. It’s the easy way out and hiring pools are increasingly filled with it.

Unprecedented levels of thinking were outsourced end-to-end: from the first interview question to the actual work after hire.

By the time I saw how many people were actively choosing not to think for themselves, the damage was almost irreparable.

It nearly forced me to shut my company down.

If I wanted to survive, there was only one path forward:

Learn how to sniff out outsourced thinking instantly and stay away from the stench.

Example 1: The Diagnostic With Buried Landmines

Everyone starts with a questionnaire. Not a vibe check, a minefield.

I plant simple, undetectable tests to see if someone’s paying attention along the entire applicant journey.

My favorite: I’ve slipped a description of our company that’s not just wrong… it’s offensively, cartoonishly wrong…into the primary source for a popular AI chat tool.

Last week alone: 61 engineer applicants.

59 used that cartoonish description in their questionnaire answers.
Eliminated.

Example 2: The First-Call Judgment Test

If you survive the diagnostic, the first interview isn’t a coffee chat.

It’s a scalpel.

I run it on four rules, disclosed upfront:

• We record.

• No ambiguity. No “that’s not what I meant.”

We pick one diagnostic answer and dig until there’s no bottom.

If you drift off-topic, I cut you off. This isn’t about being rude or controlling. It’s about effectively killing fluff.

Example: “How confidently can you identify an FTC compliance violation in an influencer post?”

If you wrote “extremely confident,” I dig until you either collapse…or leave the team speechless with a rare superpower we don’t have. That one person is why this process exists.

Hands in front of camera, off the keyboard.

Awkward, yes. But after the interview cheating that wrecked our team in 2022, I don’t apologize for it.

Discomfort by design.

The right candidates are extremely into it. The wrong ones hate it. That’s the point.

Result:

The process filters out applicants who rely on external assistance during interviews or in their work.

It identifies candidates who can sustain focused, independent thinking under pressure.

It attracts candidates who enjoy rapid-fire, respectful sparring and want their daily work to feel like lifting heavy weights at the gym, but for the brain.

Above all, it counteracts AI-driven behaviors that have quietly, rampantly reshaped hiring (at least so far). It’s iterative by design.

Lesson:

Interrogate your own mistakes. Then design safeguards ruthless enough that you’ll never repeat them.

Kaeya