Conviction Isn't Reserved for Pioneers

In Fast Company, Kai-Fu Lee, who has spent four decades building artificial intelligence, makes a case worth reading slowly. In Why taste, conviction, foresight, and courage are the traits we need in the age of AI, he argues that every industrial revolution made something expensive cheap, and this one is making cognition itself cheap. As machines get better at analysis, he names four traits that will keep humans valuable: taste, foresight, conviction and courage.

His explanation of why AI struggles with these traits is sharp. Models lean toward consensus, past data and the answer the user wants to hear. People who change things often bet against all three. Lee illustrates each trait with a famous founder, from Coco Chanel to Jeff Bezos, and he is candid that those stories carry survivorship bias. Plenty of people with the same courage were simply wrong.

Conviction Isn't a Founder's Trait

I agree with Lee's list. I want to push on where he locates it. When conviction is illustrated only through legends, most people quietly conclude it belongs to someone else. They picture a founder betting everything on a rocket launch, not a nurse who speaks up about a patient, a teacher who refuses to give up on a student, or a manager who tells the truth about a project everyone else is spinning.

Yet that everyday version is where conviction is actually built. I describe conviction as the earned capacity to act on what you believe before the outcome is certain. The word that matters most is earned. Conviction doesn't arrive with a personality type or a bank account. It grows from knowing who you are well enough to trust your own judgment, then testing that judgment in small moments until you can rely on it in large ones.

That distinction matters more in the AI era, not less. If a model can produce the consensus answer instantly, the only thing left for a person to add is a point of view the data doesn't yet support. That point of view has to come from somewhere. It comes from your experience, your values and the patterns you have learned to see because of who you are. People who have spent their careers editing themselves to fit the room have very little to draw on when the room asks for a stand.

I explored how young people build this capacity in Agency Isn't Born. It's Earned in the Rooms We Create. The same is true for adults. Conviction is practiced long before it is ever tested in public.

Borrowed Confidence Looks a Lot Like AI

Lee observes that alignment pushes models toward a safe middle. Many organizations have done the same thing to their people, long before AI. They reward the presentation that matches the plan, the forecast that matches last quarter and the opinion that matches the boss. Over time, people learn to borrow confidence from consensus instead of earning it from their own thinking.

In that sense, the risk Lee describes isn't only that machines will think like the crowd. It is that we already trained many people to, and now a machine can do it faster and cheaper. The professionals who will stand out are the ones who never fully gave up their own voice.

None of this means leaders should celebrate every contrarian bet. Shared standards, proven processes and honest data are what keep a company from mistaking stubbornness for vision. The point is balance. A good system protects a team from reckless calls. A good leader makes sure the system doesn't also silence the person who sees what the data can't show yet. I made a related point about human judgment in AI Can Learn Your Expertise. It Can't Learn Who You Are.

Practicing Conviction This Week

For professionals: the next time you reach for AI to answer a hard question, write down your own answer first. Then compare. Notice where you disagree with the model and ask why. That disagreement is often where your value lives.

Keep a short record of the calls you made before you had proof, and how they turned out. Your track record is the evidence your conviction stands on, and most people never look at it.

For leaders: in your next meeting, ask the quietest person in the room what they would do if the data were wrong. Then thank them for answering, whatever they say. Every team keeps score of how independent thinking is received.

And ask yourself which of your team's recent decisions simply followed consensus. If the answer is most of them, a model could have made them too.

Lee is right that AI will make intelligence abundant. What it can't make abundant is people who know themselves well enough to stand behind a belief before the world agrees.

Conviction isn't reserved for pioneers. It belongs to anyone willing to know who they are and act on it before the proof arrives.

Want to explore these ideas further? Learn more about my work on leadership, identity and conviction at www.theglennllopis.com.