Skip to content
Work Tech Weekly
Matt Poepsel

The Workplace Psychology Your AI Tools Are Missing

Most AI conversations right now are about speed.

How fast can we deploy it? How quickly can we cut costs? How soon can we see returns? But inside most organizations, the people side of that equation is barely getting a fraction of the attention. Executives are pushing hard. Employees are overwhelmed. And managers are stuck in the middle, expected to hold everything together with almost no support.

Matt Poepsel, Vice President and Godfather of Talent Optimization at Predictive Index, has a simple explanation for why. The quality of advice you get from any AI tool is only as good as the context you feed it — and most organizations are missing the most important context of all. Not the data in their CRM or their HRIS. The data about how their people are actually wired.

I sat down with Matt at Transform 2026 in Las Vegas — outside on the deck at the Wynn, overlooking a golf course neither of us had time to play — to talk about why AI deployments keep falling short, what managers actually need right now, and what behavioral science has to do with any of it.

At his session earlier that week, Matt had asked the room a simple question: how much of the energy in a typical AI deployment goes toward technology versus people? The audience didn’t hesitate. Eighty to ninety percent technology. Ten to twenty percent people.

Every business problem is a people problem. Most organizations are still treating AI like it’s a technology problem.

The Workplace Psychology Your AI Tools Are Missing
  33 min
The Workplace Psychology Your AI Tools Are Missing
Work Tech Weekly
Play

The Manager Caught in the Middle

There’s a reason managers are struggling right now, and it’s not that they’re bad at their jobs.

“Executives that wanna go fast and change things, and an employee base that’s already overwhelmed,” Matt says. “It’s never an easy environment. Right now it’s particularly challenging.”

Managers are absorbing tension from both directions simultaneously. The C-suite wants AI deployed yesterday. Employees are watching companies lay off thousands in the name of efficiency and losing trust in real time. And the manager in the middle is expected to hold the team together, hit the numbers, and lead people through uncertainty — usually without training, without data, and without a clear picture of how the people on their team are actually wired.

“The average manager doesn’t get a lot of training anymore when it comes to how people are wired, really what their natural preferences are,” Matt says.

That gap has always existed. What’s changed is the stakes. When a manager doesn’t understand how someone on their team responds to a directive leadership style, or to the kind of rapid change AI deployment demands, they lose that person’s trust before the tool is ever rolled out. The deployment fails not because the technology doesn’t work. Because the conditions for it to work were never built.

Predictive Index has spent decades working on what Matt calls the consumability problem — getting behavioral insights into the hands of a line manager who doesn’t have a PhD and doesn’t have time to read a static report. The answer, increasingly, is that the insights have to live in the flow of work. Not in a report you pull up twice a year. In the meeting you’re about to have. The performance review you’re drafting. The conversation where something just went sideways.

Why Your AI Tool Doesn’t Know Your People

Here’s the version of this problem most organizations are already trying to solve: give managers access to AI tools and let them ask better questions.

The problem is what they’re feeding those tools.

“If I’m feeding the AI my slanted view of the situation and it’s responding, give me advice, that’s not good advice,” Matt says. “I don’t know scientifically the nature of the people who are around me.

Generic AI doesn’t know how your people are wired. It knows what you tell it — which is usually one manager’s read of a situation, filtered through their own blind spots. The advice it generates reflects that. It’s guidance built around a description of a problem, not the actual dynamics underneath it.

Behavioral data changes the equation. When an AI system knows not just someone’s job title but their drives, their needs, their communication preferences, how they respond to change — the advice it generates is specific instead of generic.

Matt puts it plainly: “Help me word this email to Bob. Okay, well you have to tell me a little bit about Bob. But in PI’s world, no you don’t. Because we already know about Bob.”

That’s the gap most organizations haven’t closed. They have the tools. They often have the data. What’s missing is connecting the two — putting behavioral context into the systems where managers actually work, so the advice they get back reflects the real team dynamics, not a one-sided account of them.

What HR Has to Do Differently — Now

The urgency in this conversation isn’t abstract.

Matt is direct about the stakes for HR specifically. The field has long struggled to translate its value into business terms — to show up to the room with data instead of process expertise, to make the argument in language that lands with a CFO or a CEO.

That used to be a missed opportunity. Now it’s something else.

“Not being able to articulate HR’s value in business terms was unfortunate. It was a little bit of a missed opportunity. Today I say it’s career limiting,” Matt says

AI is automating parts of the HR function itself. The organizations that come out of this with strong people practices will be the ones where HR learned to show up differently — with behavioral data, with a clear argument for why understanding how people are wired matters more than any tool they’re deploying. Here’s what we know about this team. Here’s how they’re likely to respond. Here’s what we need to do to get the outcome we’re after.

The window for that shift is narrower than most HR leaders realize. There’s a perfect storm bearing down on the workforce right now — AI uncertainty holding back hiring, Boomers and Gen Xers staying in the workforce longer than expected, a promotion pipeline that’s backed up, entry-level workers trying to get in the door at exactly the wrong moment. The people problems are compounding. The organizations equipped to navigate them will be the ones where HR already made the pivot.

“The opportunity to really be more of a translator of people terms into business reality, that opportunity is very much there for HR,” Matt says. “But it’s going to take something very different than what historically the field has brought to the table.”

The stat Matt shared at Transform — 80 to 90 percent of AI deployment energy going toward technology, almost nothing toward people — is a kind of X-ray. It shows exactly where the investment is and exactly where the risk lives.

Behavioral science doesn’t get weaker when AI enters the picture. It gets more important. The better the context you feed an AI system, the better the guidance it gives back. And most of that context isn’t in your CRM or your HRIS. It’s in the people data that organizations have already collected and quietly set aside.

The tools exist. The data is often already there. What’s missing in too many organizations is the willingness to use it.

Let's Talk