Across many interviews with AI experts and executives, one specific follow-up question has proven unusually reliable at distinguishing a source's genuinely considered view from a polished talking point delivered on autopilot: asking directly what would change their mind.
Why this question works so well
A rehearsed talking point is built to be delivered confidently and repeatedly, and most sources haven't prepared a similarly polished answer for what evidence would actually change their stated position - a question that requires genuine, in-the-moment reasoning rather than a recitation.
What the answers tend to reveal
A source with a genuinely considered position can usually articulate, at least roughly, what evidence would update their view, even if they don't expect that evidence to materialize soon. A source reciting an institutional talking point often struggles meaningfully with this question, because the position wasn't built through a falsifiable reasoning process in the first place.
How we use this in practice
We ask this question routinely in interviews, and the quality and specificity of the answer often shapes which quotes from a conversation we consider strong enough to feature prominently versus which ones we treat as more generic supporting color.
Why we think this matters for readers
A quote from someone who's genuinely reasoned through their position, including what would change it, carries more informational value than an equally confident quote reciting an unfalsifiable institutional line, even when both read as similarly authoritative on the page.