"They just gave people tools and said figure it out — and that's not going to happen." — Susan Guest

The pattern is everywhere: AI declared a priority, tools deployed, upskilling on the strategic plan. And the work still isn't moving.
The problem isn't awareness or budget. Organizations are treating a transformation challenge like a technology rollout — and the people side keeps getting left behind.

Who helped move the work forward

This live interactive discussion brought together people leaders, L&D strategists, organizational development practitioners, and AI adoption specialists from organizations across healthcare, insurance, financial services, manufacturing, and consulting. Facilitating the work: Adrienne Guerrero (Positive Delta), Vilson Simon (Ágora & LAK Group), Marty Murrillo (Precisely), and Mike Hoyt (ELE), with the Baryons AI Mentor participating as a live synthesis partner.

Participants worked through role-specific AI prompts — taking on the perspectives of senior leaders, frontline managers, and HR business partners — to build execution-ready action plans for unsticking upskilling in their own organizations.

The signals and the incentives aren't aligned

Everyone in the room could name it. Leaders say upskilling matters. The reward system says hit your numbers. And the manager layer — the people who actually control whether employees get time to learn — is stuck in the middle with no incentive to open the door.

"If you can't get through the door to get coffee, you can just look in the side window and watch the machine — does you no benefit." — Anton Maletich, Alter Domus

The answer the group kept returning to: give managers explicit permission, then tie it to how you evaluate and reward them. One without the other doesn't move.

"Deploy the tool" is not a transformation strategy

This was the sharpest shared signal of the discussion — and it surfaced across every breakout group.

Organizations are procuring AI as a SaaS product, tracking token usage, measuring productivity gains, and calling it transformation. They're missing the point.

"This isn't about technology — this is about transformation, rethinking business processes, workflows." — Marty Murrillo, Precisely

Karen Waterlander named what that gap looks like on the ground: people know the tools exist, they know they're embedded in some systems, but there's no clear direction on when, how, or why to use them. The result is people "floundering."

The reframe that emerged: stop measuring whether people are using AI. Start asking what they're doing with it.

"Are they reimagining their workflows? Are they reinventing new lines of business? Not just are they using it to increase productivity." — Marty Murrillo, Precisely

Adrienne Guerrero shared a client example: hackathons built around AI process refinement surfaced workflow problems that had nothing to do with AI. Some of the biggest wins came from simply stopping things nobody had questioned in years.

The metrics aren't measuring what matters

Most organizations are tracking consumption — LMS completions, hours logged, token usage. None of that tells you whether work actually changed.
Vilson Simon framed the stakes: 39% of worker core skills will be transformed or obsolete by 2030, and 63% of employers already name skills gaps as the top barrier to business transformation. If that's true, the measurement has to match.
Anton Maletich's group moved toward more useful signals: learning hours alongside client satisfaction scores, manager stress indicators, and lines-of-business comparisons to see where investment is actually shifting outcomes.

---

What to try next

Run all three role prompts — then synthesize. Senior leader, frontline manager, HR business partner. Run them independently, then feed the outputs into a single thread. If the three contradict each other, you've found where your org is misaligned.

Give managers the permission in writing — and tie it to their evaluation. One to two hours per week, explicit from the top. Without the incentive attached, the mandate stays theoretical.

Use AI to focus learning time, not just deliver content. Before the learning hour, run a quick AI diagnostic on where to focus. Anton Maletich's team is piloting Outlook MCP integration with Claude to surface development signals directly from work context.