Over the past few months, I've noticed something changing in technical discussions. It used to be:
"I think we should do X because..."
Then came arguments, tradeoffs, benchmarks, past experiences, and sometimes healthy disagreement. Now I'm increasingly hearing:
"I asked ChatGPT (or Claude), and it recommended X."
Or even:
"The AI says this is the best approach."
I'm not questioning the usefulness of AI. I use it every day.
What concerns me is when AI stops being another input and starts becoming the argument itself. LLMs can be incredibly persuasive, they often sound confident, even when the answer depends heavily on context. The way a question is phrased can also influence the response, making it surprisingly easy to receive an answer that supports an existing opinion.
And that makes me wonder... How are other teams handling this?
Have you ever been in a meeting where "the AI recommended it" was treated as the end of the discussion?
If you're in a leadership position, how do you balance AI recommendations with the experience and judgment of your team? And if you've disagreed with an AI recommendation, how did you approach that conversation?
I'm genuinely curious because this feels like a challenge many teams may be facing as AI becomes part of everyday development.
Where do you draw the line between using AI as a tool and treating it as an authority?