Everyone on the panel is reading from the same script
By Precious Ebere-Chinonso Obi
I’ve lost count of how many panels I’ve sat on this year and lately, somehow I’ve started to notice the same pattern: we’re all saying versions of the same thing. I do not think it is because we all agree but because we asked the same machine.
More panelists than ever are leaning on AI to generate their talking points before a discussion and the inconvenient truth is that most of us are leaning on the same model, asking it more or less the same question.
READ ALSO: “Hold the government accountable” is not enough: What Nigerians must actually do
The result isn’t four perspectives in conversation with each other. It is one perspective, recited four times in four voices. The disagreement that used to make panels worth attending, the moment where someone pushes back and the room leans in is quietly disappearing.
What’s left is consensus dressed up as discussion and it is really bothersome to me that I had to change how I prepare.
Using AI as a floor, not a ceiling
Here’s the rule I’ve settled on: I still use AI before I speak but I use it for exactly one purpose to find out what everyone else is probably going to say.
If I ask a model to outline talking points on a topic, I’m not looking at its answer as my contribution. I’m looking at it as the consensus. It’s the average of everything that’s already been written and said, smoothed into the most statistically likely response. That’s useful information but it’s the floor of the conversation, not the ceiling.
Once I know what the “obvious” answer looks like, the real work starts. I ask myself: what’s missing from this? What would the room naturally overlook? Who isn’t represented in this synthesis, and what would they say instead? That question: what’s the missing link is where my actual thinking begins. It’s not a prompt I can outsource, because by definition it’s looking for what the machine didn’t surface.
It’s a small discipline, but it’s changed what I bring into a room. I’m no longer trying to be the most polished voice repeating the consensus. I’m trying to be the one who noticed what the consensus left out.
Why this matters more than it used to
I don’t think this is a minor professional quirk. If most public commentary, most panel discourse, most “expert opinion” is quietly converging on the same AI-generated starting point, then the diversity of thought we assume exists in public conversation may be thinner than it looks. We can have a room full of different job titles, different backgrounds, different industries and still walk away having heard one idea, repeated.
The fix isn’t to stop using AI before high-stakes conversations. It’s pretty good at telling you what everyone already thinks. The fix is treating that as the starting line, not the finish line, and doing the slower, harder work of asking what it left out.
I was reminded of this discipline recently reading a piece by Michelle Odemwingie, CEO of the Achievement Network, who describes a near-identical moment on a panel of her own realizing mid-discussion that another speaker was echoing the same AI-generated prep she’d used that morning. Her essay frames the underlying choice as one between AI that extends human judgment and AI that quietly substitutes for it. It’s a useful frame, and it’s worth reading in full.
My version of the same instinct is narrower and more practical: before I speak, I ask not what the machine said, but what it didn’t.
That question is the only thing in the room that’s actually mine.
- Precious Ebere-Chinonso Obi, CEO of Do Take Action, is an independent consultant on edtech, climate change, public policy, and women’s procurement empowerment
