Unstructured Interviews Meet AI Note-Takers
Let’s start with Reddit. If you ever wanted to feel equal parts entertained and alarmed about the state of interviewing, take 15 minutes and explore a hiring thread. If you don’t have 15 minutes, here are just a few questions candidates report being asked:
“If you had to choose between being an elephant the size of a rat or a rat the size of an elephant, which would you choose?”
“What Disney princess would you be?”
“How many phone books are in New York City?”
“If you were leaving a nightclub with Prince Harry, how would you escape the paparazzi?”
“Would you rather be covered in spiders or eat one alive?”
“Do you plan on having children in the near future?”
If you’re a candidate, it’s a horror show. If you’re an HR leader, it’s disappointing. If you’re an executive team trying to improve performance, it’s…concerning.
For years, we’ve rolled our eyes at these types of questions. They were annoying, silly, sometimes misguided attempts at being clever. Many were okay pretending it wasn’t happening or that it wasn’t harmful. Now they’re being documented, and the damage they have been causing is coming into view.
The Chaos Was Always There
Interviews have always been a gray area. Historically, they’ve disappeared when the conversation ended. Someone jotted down a few notes, general impressions were shared, and a decision was made.
Despite the emphasis on hiring top talent, interviews have been a high-stakes step in the process, often the decisive moment, yet they are also loosely structured, highly subjective, and inconsistently executed. AI notetakers are capturing and publishing these inconsistencies.
Another Reddit thread featured hiring managers sharing their favorite interview tactics. One proudly described taking the candidate’s resume away and grilling them on dates and accomplishments to “see if they made anything up.” Another asked candidates their favorite book to determine whether they were “well read” enough to be intellectually impressive. This manager went on to state that if a candidate answered “50 Shades of Grey,” that told him everything he needed to know about their workplace appropriateness. And then there were the paper airplane exercises, timed speeches, and other curveballs designed to throw candidates off balance so they could evaluate the “real person.”
These personal experiments don’t improve hiring; they reveal a lack of shared success criteria, structured evaluation standards, and validated data tied to job requirements.
Just “Vibe Checking”
Another Reddit thread featured hiring managers sharing their favorite interview tactics.
One proudly described taking the candidate’s resume away and grilling them on dates and accomplishments to “see if they made anything up.” Another asked candidates their favorite book to determine whether they were “well read” enough to be intellectually impressive. This manager went on to state that if a candidate answered “50 Shades of Grey,” that told him everything he needed to know about their workplace appropriateness.
One Reddit user summarized the strange questions with, “They’re about vibe checks.”
These personal experiments don’t improve hiring. Imagine reading any of these questions in a transcript six months later during a hiring dispute. Or seeing it quoted in a social media post. Or trying to explain how it relates to job performance in front of legal counsel.
Whether hiring managers are trying to build rapport, fill awkward silences, or are simply improvising, they are creating risk, intentionally or unintentionally. Without structure, hiring becomes a collection of individual experiments and philosophies.
AI note takers weren’t introduced to fix interviews. They were introduced to reduce note-taking and improve efficiency. But the consequences are unavoidable. Unstructured interviews have always been weaker predictors of performance. Research has shown that for decades. But they persisted because they felt human, flexible, and relational.
When one candidate is asked about Disney princesses, and another is grilled about their reading habits, that’s no longer stylistic variation. When one interviewer probes deeply, and another barely scratches the surface, that creates uneven evaluation. When questions move into legally questionable territory, they don’t disappear anymore.
The Deeper Issue
In many organizations, interviews are set up to do everything. They are expected to:
Discover a candidate’s capabilities
Assess motivation
Predict performance
Evaluate fit
Identify derailers
That’s an enormous burden for a conversation that varies from interviewer to interviewer. The inconsistency becomes consequential, and this is often where HR leaders start to feel the tension.
Business leaders want autonomy and speed.
Executives want better performance outcomes.
Legal wants defensibility.
And now, documentation makes interview quality visible in ways it never was before. The conversation needs to shift from “Did we hire the right person?” to “Is our hiring process designed to hire the right person for the job?”
Mature hiring processes:
Define what success looks like
Tie interview questions to those criteria
Use structured rating scales
Compare candidates against shared standards
Train interviewers on how to evaluate data
What If
If one of your interviews were posted on Reddit tomorrow, would you feel confident explaining how each question ties to performance in the role? If an AI transcript were reviewed by legal, the board, or a candidate, would the logic behind your process be clear?
Transparency isn’t going away. And unstructured interviews aren’t just quirky anymore. The problem isn’t AI. The problem is expecting informal interviews to carry the weight of a strategic hiring decision.