For decades, workplace learning followed a relatively structured path. Employees completed courses, absorbed concepts and only then applied what they had learned on the job.

That model assumed work was predictable and linear. Today, it rarely is.

Instead of studying for weeks in advance, some are starting with intent of a clear outcome they want to create and are relying on artificial intelligence (AI) to fill gaps between idea and execution by generating options, highlighting trade-offs and guiding users through early decisions. As a result, the way we work has changed. Teams progress through rapid cycles of review and refinement with AI acting as a real collaborator. If this is how work now happens, learning must evolve to match it.

The Rise of “Vibe Learning”

We can think of this new model as “vibe learning:” employees learn through building, experimenting and refining ideas with AI assistance.

Rather than separating training from work, AI-assisted learning blends them into a single unfolding process. Employees begin with a project or problem, lean on AI for structured support and gain new skills as they move toward a tangible outcome.

Traditional curricula and structured learning paths still have an important role — they create consistency, shared language and foundational knowledge. But taken alone, they assume a degree of predictability that no longer exists.

Work today is dynamic: projects evolve quickly, cross functional collaboration is the norm and employees are increasingly expected to stretch beyond fixed roles. Learning systems must be able to flex with that reality.

That means complementing formal curricula with approaches that can:

• Recommend skills based on live projects.
• Support stretch assignments across teams.
• Personalize learning around real tasks and goals.
• Measure progress through outputs and contributions rather than course completion.

In other words, the focus shifts from course consumption to skill momentum.

Redesigning an Onboarding Program Using Vibe Learning

One of the most powerful aspects of AI-assisted learning is its ability to reduce the friction of starting. Consider the following illustrative example:

An HR team is tasked with overhauling its onboarding process to reduce time-to-productivity for new hires. Under a traditional model, the team would enroll in a course on instructional design, complete it, and then begin building the new program weeks later.

Instead, the team starts with the problem. Using AI, an HR generalist with no formal design background begins prototyping a new onboarding workflow — drafting competency checklists, generating role-specific learning paths and mapping content to 30/60/90-day milestones. AI doesn’t just produce outputs; it explains the rationale behind each recommendation, surfaces alternatives and flags where assumptions may not hold across regions or roles.

A small, cross-functional group — an HR business partner (HRBP), a recruiter and a people analytics lead — works together through the build, pressure-testing the prototype with their own expertise. Each iteration sharpens both the onboarding program and the team’s understanding of instructional design, data-informed decision-making and change management. Skills are built through the process rather than developed in advance of the work.

The result: a stronger onboarding experience built faster and, just as important, an HR team with new capabilities they didn’t have before they started.

How to Make AI-Assisted Learning Work

Of course, adopting a build-first learning model does not mean abandoning structure or instructional design. Organizations still need guardrails to ensure experimentation aligns with business goals.

For learning and development (L&D) leaders, the challenge is to create environments where this type of learning can thrive while maintaining strategic focus. Practical steps include:

Embed Learning in Real Projects.

Encourage teams to treat new initiatives as learning opportunities, pairing AI tools with structured reflection and feedback. Start with intent, asking your team what they’re trying to accomplish and use that real work as the container for learning, rather than assigning training in advance.

For example, a marketing team might start with a campaign they need to launch and use AI as a collaborator to generate ideas, test messaging and refine content, learning necessary skills as they move toward a finished deliverable, not a hypothetical exercise.

Integrate AI Agents Into Training Programs.

Instead of relying solely on courses, allow learners to prototype ideas and experiment with guided AI support. This means pairing foundational instruction with hands-on practice, where AI acts less like an answer engine and more like a coach — prompting reflection, surfacing alternatives and helping learners understand why one approach works better than another.

Help employees learn how to use AI to explore different approaches, understand the logic behind outputs and continuously improve their work through iteration. In some cases, this support shows up directly in the flow of work, offering real-time guidance as employees prepare for or navigate high‑stakes tasks and conversations.

Recognize Outcomes as Evidence of Learning.

Shift evaluation from course completion toward tangible outputs, prototypes or problem-solving contributions. When learning is embedded in work, progress becomes visible through clearer decisions or more effective workflows rather than attendance or completion rates alone.

In practice, this means assessing how employees apply and refine their work overtime — for example, how a team improves a workflow or sharpens campaign messaging as they work toward a better outcome. These changes provide leaders with clearer evidence of skill development and make learning easier to connect to business impact.

Encourage Experimentation Over Perfection.

Learning accelerates when employees feel safe exploring ideas and iterating quickly. Creating space for experimentation helps reinforce this behavior.

That safety doesn’t mean removing standards but setting clear expectations, like treating AI output as a draft, defining who reviews it and clarifying accountability, so experimentation leads to better judgment.

When these elements are in place, learning becomes a byproduct of meaningful work rather than a prerequisite for it.

Rethinking Learning for the AI Era

AI has already transformed how work gets done. The next step is allowing it to reshape how we help people grow — so learning keeps pace with the reality of modern work.

For organizations just beginning to shift toward AI-assisted learning, the starting point doesn’t need to be a full transformation. Piloting this approach within a single team or project by introducing AI-assisted build sessions, defining guardrails and measuring outcomes, can provide a practical way to test and scale the model over time.