On AI and Decision-Making

Much of the discussion around AI assumes that better answers will lead to better decisions.

The assumption seems reasonable.

If information becomes easier to access, faster to analyze, and cheaper to obtain, why wouldn't people rely on it?

The assumption, however, overlooks something important about how humans make decisions.

Years ago, while conducting research at NerdWallet, I spent time studying how people make major financial decisions. Whether they were buying a home, refinancing a mortgage, selecting insurance, or making investment decisions, one pattern appeared repeatedly.

People rarely relied on a single source of information.

They researched online.

They consulted experts.

They read reviews.

They asked friends.

They talked to family members.

They looked for people who had gone through similar experiences.

What stood out wasn't the amount of information they gathered.

It was the variety.

People weren't simply collecting facts.

They were building confidence.

I've seen the same behavior in healthcare.

A patient receives a diagnosis and seeks a second opinion.

Not necessarily because they distrust the first physician.

Because the decision carries weight.

The same thing happens when people choose a college, change careers, make a large purchase, or evaluate a significant financial commitment.

The more consequential the decision, the more likely people are to seek multiple perspectives.

This is what makes some predictions about AI particularly interesting.

Many assume AI will eventually become the primary source of decision support.

That may happen for low-stakes decisions.

But high-stakes decisions often involve more than information.

They involve uncertainty.

Trust.

Risk.

Responsibility.

People want to know what others think.

They want reassurance.

They want to hear from someone who has been through it before.

They want confidence that they're not overlooking something important.

AI may become one of those inputs.

Perhaps a very powerful one.

But that doesn't necessarily mean it replaces the others.

In fact, as information becomes easier to access, trust may become more important rather than less.

The challenge shifts from finding information to evaluating it.

Who should I believe?

Which recommendation matters most?

What perspective am I missing?

The future of decision-making may not be AI versus human judgment.

It may be AI joining a process that has always relied on multiple perspectives.

Because when the stakes are high, most people aren't looking for a single answer.

They're looking for enough confidence to make a decision.