Artificial intelligence gurus give good advice when they tell me: Don’t treat AI like a smarter search engine; treat it as a collaborative thought partner. But how good a job am I doing with human thought partners?

Not as good as I could.

Long before ChatGPT brought AI to the masses in late 2022, I’d been underusing the intelligence of nearby humans. Why don’t I pause more often to ask for a colleague’s perspective, and be truly interested in her answer? I’m busy, distracted, confident (read: certain). And now I’ve got a new alternative in chatbots that, unless I’m careful, might tell me what I want to hear without any of the messy friction of human exchange.

That messy friction—that’s the opportunity.

Each colleague carries a lifetime of idiosyncratic experiences, beliefs, and context—far richer than any AI model can simulate. Over decades the gifts of colleagues’ advice have saved me from bad trouble and nudged me into good, exposed weak thinking and nurtured strong, and prompted more instances of “I hadn’t thought of that” than I can count.

Once I asked for thoughts about an advisory panel I was assembling. “The nominees are strong, but they all like you,” my colleague said. “Include someone who doesn’t.”

What if I’d pursued more human thought partners, especially those more likely to differ?

Smarter people than I have given this careful thought. Edgar Schein taught us to build trust by switching from “telling” to “asking.” Amy Edmondson reminds us that if we’re going to ask a colleague for her opinion, we need to make it psychologically safe to respond.

While we’re exploring their insights, we can act now to grow human-to-human thought partnering by asking three questions of others:

  • What do you think?
  • What am I missing?
  • What would you do?

What would you do to increase human-to-human thought partnership? Share your thoughts in comments.