AI is rewriting how work gets done—but most organizations aren’t designed for it.
Despite historic investment in AI, productivity gains have been modest because companies are automating tasks inside operating models that were never built for AI agents. Job architectures are outdated. Processes are fragmented. Workflows don’t reflect reality. And employees are stuck in roles that no longer match the work.
Josh Bersin’s new report, Dynamic Work Design: The Key to AI Transformation, provides the clearest explanation yet for why AI progress stalls—and what leading organizations are doing to fix it.
The Core Insight: AI Can’t Transform Your Business Until You Transform the Work
Most companies begin with technology—rolling out copilots, assistants, and tools that scatter across functions. The result? Incremental improvements at best.
Bersin’s research shows that sustainable productivity requires a shift from job-based design to outcome-based work design:
- Start with the business problems
- Redesign work around outcomes and accountabilities
- Identify tasks that can be automated, augmented, or redesigned
- Build roles that combine human strengths with AI capabilities
- Create a continuously evolving operating model
This approach is what Bersin calls Dynamic Work Design—a new organizational muscle for the AI era.
The Four Stages of AI Work Transformation
The report outlines four stages companies move through as they modernize work:
- AI Assistants: Task automation for individuals (15–30% improvement)
- AI Agents: Workflow-level automation (30–50%)
- AI Multifunctional Agents: Human + AI collaboration redesigns roles (100–200%)
- AI Autonomous Agents: End-to-end process transformation (300%+)
Only Stages 3 and 4 deliver true transformation.
And both require deep visibility into how work actually happens.
Why Work Intelligence Is Now Essential
The report highlights a critical gap: most enterprises can’t redesign work because they lack work intelligence—the connective tissue that maps:
- Tasks and subtasks
- Workflows and processes
- Roles and activities
- Required skills
- AI opportunities
This layer has never existed in HCM or job architecture systems.
It is emerging now as a new category of platform: Work Intelligence.
How Leading Enterprises Are Using Work Intelligence Today
WPP
Used work intelligence to consolidate 55,000 job titles into 600 roles, identify automation and augmentation opportunities, and redesign work for higher-value activities.
Micron
Shifted from static job architecture to a dynamic model that maps tasks, skills, and capabilities in real time—improving resource allocation for engineering and manufacturing teams.
These cases demonstrate the power of pairing dynamic work design with task-level insights.
Why This Matters for Leaders
For CEOs:
A blueprint for unlocking productivity and building an AI-ready operating model.
For CIOs:
A strategic approach to deploying AI agents where they create real value.
For CHROs:
A path to become the architect of future work and talent.
The Bottom Line
AI won’t deliver its ROI until work itself is redesigned.
- Dynamic Work Design offers the framework.
- Work Intelligence provides the data and platform to make it real.
