The Collapse of the Competence Framework
For nearly three decades, I have stood at the front of training rooms on every continent. I have coached executives in Paris, guided leaders under the shadow of the Taj Mahal, and taught new managers in Dubai’s boardrooms. I have seen thousands of leaders wrestle with the same recurring questions: Which skills matter most? And, how do we prepare our people?
Today, those questions are collapsing under the weight of something new. AI is not simply helping us perform skills. In many tasks, it has actually become THE skill.
From drafting contracts to coding software, analyzing data, and generating leadership reports, I see AI deliver work to a standard that once required years of human practice. For HR leaders, this demands a new conversation. It is no longer about which skills to add to the catalog, but about what remains uniquely human when “competence” itself is shifting to the machine.
What the Evidence Tells Us
This is not my opinion alone. The research is clear.
- In one field study of a U.S. customer-support operation, AI guidance lifted productivity by 14% overall, with the greatest benefit for novice agents. Inexperienced employees gained the equivalent of months of learning in weeks (Brynjolfsson et al., 2023).
- Controlled experiments in knowledge work found participants completed tasks 40% faster and with higher quality when using AI assistance (Noy & Zhang, 2023).
- Global surveys from IBM, McKinsey, and the OECD show executives now expect a third to nearly half of their workforce to reskill within three years due to AI reshaping tasks.
As someone who has built leadership academies, delivered executive education, and advised CHROs across industries, I know how quickly the ground can shift. This time, however, it is not simply a change in demand for certain skills. It is the redefinition of “competence” itself.
What is Really Dying?
It is not learning. It is not development. What is dying is our old definition of skill: the idea that skill is a memorized routine executed faster with practice. Machines can already replicate those routines, learn them, and adapt them at a pace no human can match.
What survives, and grows more valuable, are human capacities that AI cannot replicate:
- Defining the right problem before solving it.
- Making principles-based trade-offs under pressure.
- Safeguarding values and ethics when technology accelerates.
- Creating belonging, trust, and meaning in human communities.
In every leadership session I deliver, I tell executives this: If you think AI erases the role of people, you are approaching this from the wrong angle. Yes, this could be true, but only if we are unprepared. The real issue is whether you are willing (and able) to redesign work so that human judgment and machine execution are woven together intelligently.
The Competence Stack: A Framework I Often Use with Leaders
To help organizations adapt, I use a model I call the Competence Stack. It reframes capability from a list of skills into a layered system where people and AI complement each other.
Layer 1: AI Fluency and Data Literacy: This is the new workplace grammar. Every employee must know how to frame prompts, verify results, respect privacy, and understand tool limits.
Layer 2: Meta-Capabilities: These are the higher-order abilities that distinguish human oversight from mechanical execution. They include problem framing, ethical reasoning, cross-functional sense-making, and authentic communication.
Layer 3: Domain and Context: Here lie the elements only people can judge. Regulations, customer promises, cultural memory, and organizational values. AI can propose. Leaders must decide.
When I walk executives through this stack, the insight is immediate: stop treating skills as static checklists. Start treating “competence” as a living system that shifts as AI takes on baseline tasks.
Once leaders grasp the stack, the next question becomes obvious: what must HR rewire today?
What HR Must Redesign Now
1. Job Architecture: Roles built on static job descriptions are already outdated. Recode them into outcome maps that assign responsibility as human-led, AI-led, or shared.
2. Metrics: For years, I have challenged boards that track “training hours” as if they measure performance. Replace vanity measures with signals that matter: time to competence with AI, error rates with AI in the loop, explainability of automated decisions, and employee trust in AI processes.
3. Learning Supply: Annual catalogs will not keep up. Build a skills graph that draws live signals from hiring, AI usage data, and performance. Push learning into the flow of work, not in occasional workshops.
4. Governance: Every HR process touched by AI must go through a simple, visible review: what is the purpose, what data is used, what bias checks are in place, what appeal routes exist, and how employees are informed. I stress this in every boardroom: trust is not built by promises. It is built by transparency.
The 5 Traps I See Too Many Leaders Fall Into
- Buying platforms before mapping tasks.
- Freezing competency models that expire before the ink dries.
- Claiming “human in the loop” oversight without ensuring the human understands the model.
- Rewarding effort over outcome. Recognizing those who work harder, not those who build smarter systems (i.e those who redesign).
- Training features, not judgment. Running tool demos instead of practice with real cases and real data.
I tell executives bluntly: AI does not expose the weakness of technology. It exposes the weakness of leadership.
A 12-Month Blueprint I Recommend to Every CHRO
Quarter 1: See The Work: Run a task census. Tag work as human-led, AI-led, or shared. Publish a guardrail list of human-only tasks.
Quarter 2: Recode Roles: Rebuild roles as outcome maps. Assign decision rights and risk thresholds. Realign incentives.
Quarter 3: Build the Stack: Roll out AI fluency sprints. Launch a skills graph. Fund targeted academies to train judgment, not features.
Quarter 4: Govern and Scale: Institutionalize AI reviews. Embed appeal routes. Update progression frameworks to value oversight and design. Present results to the board.
What Great Leadership Sounds Like
In my work with global HR leaders, I have learned employees do not expect certainty. They expect clarity. Say three things with precision:
- What AI will do here.
- What AI will never be allowed to do.
- What I need from you to make the system safe, fair, and valuable.
The leaders who can articulate this, and prove it with action, will not only survive the AI era. They will set the cultural tone of the organizations of tomorrow.
The Road Ahead
I have built my career on training leaders to adapt to disruption. From the first rise of digital HR platforms to today’s algorithmic agents, I have seen our profession repeatedly tested. The story of skills is not ending. It is being rewritten.
“Competence” is no longer a catalog. It is a stack where machines execute, humans interpret, and leaders define what good looks like. Those who cling to yesterday’s definitions will be left behind. Those who treat competence as living, evolving, and co-created will lead the reinvention of work itself.
Food for Thought
- Which tasks in your organization are ready to be AI-led, and what oversight will they require?
- Which tasks must remain human-only, and why?
- What three metrics would persuade your board that people plus AI outperform either alone?
- How will you raise AI fluency across every role in the enterprise?
References
- Brynjolfsson, E., Li, D., Raymond, L. R. (2023). Generative AI at Work: Evidence from a Call Center. NBER Working Paper 31161.
- Noy, S., Zhang, W. (2023). Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence. Working paper.
- OECD. (2023). Employment Outlook 2023: Artificial Intelligence and the Labour Market.
- World Economic Forum. (2023). Future of Jobs Report 2023.
- Microsoft. (2024). Work Trend Index: AI at Work Is Here.
- IBM Institute for Business Value. (2023). Augmented Work for an Automated, AI-Driven World.
- McKinsey Global Institute. (2023). Generative AI and the Future of Work in America.
- LinkedIn. (2024). Workplace Learning Report.
First published on LinkedIn.