In a recent study at MIT, the author found that highly skilled employees benefitted the most from AI-enabled support. These findings contrast earlier studies, showing that more junior or less-skilled workers benefit the most. The MIT study also found that highly skilled workers enjoyed their jobs less because AI was involved. They shared that AI takes away some of the creativity humans usually bring to the process.
The challenge with that perspective is that we are judging tomorrow’s productivity by today’s standards. From understanding what it takes to become an expert to redefining our understanding of work, we need big thinking and practical approaches to prepare for AI’s impact in the workplace. So, let’s explore!
Becoming (and Remaining) an Expert in an AI-Enabled World
We are seeing a period in which we have acquired skills (before AI’s influence on work). You learn a trade, go to college, conduct research, or simply learn by doing. But now, AI enters the picture, taking over more and more mundane tasks that we have been learning and doing ourselves.
What does it mean to be an expert? And what does it take to become (and remain) one?
The latter question is especially relevant for junior team members looking to develop their careers. Canadian author Malcolm Gladwell estimated that it takes an average of 10,000 hours (or the equivalent of 5 years of full-time employment) for a novice to become an expert in a subject. Others estimate that number to be 20,000-25,000 hours (10-12 years). On this journey, we acquire skills from novice to expert level in five stages (see below).
Dreyfus, Stuart E. (2024). The Five-Stage Model of Adult Skill Acquisition. https://journals.sagepub.com/doi/10.1177
However, according to research by the World Economic Forum, 44% of our knowledge and skills will become less relevant by 2027, in part due to AI. We might need to know the fundamentals, but we are no longer expected to apply them all (or in the same way).
Think of advanced calculus and how the calculator has reshaped the process. We still need to know how to create differential equations, but we can use a new tool (calculator) to compute them for us. Similarly, learning how to work with the AI tool is becoming a more in-demand skill than completing the task yourself.
Typical expert learning curves and assumed AI-supported learning
AI boosts experts’ productivity and quickly raises the outcomes for novices. The latter is likely a “productivity jump.” Novices still need to gain practical experience, though, to develop practical expertise and assess (AI-generated) information that informs their decision-making in addition to achieving productivity gains. It means that AI-supported novices might achieve that expert-level proficiency faster. Until they do, they face an “experience gap” relative to current experts. Productivity alone will not make one an expert.
While AI shortens the learning curve for novices, they still need thousands of hours of practice to become experts.
Acquiring these skills is only possible if we adjust our mindset and redefine what work and working actually mean and how we measure success.
Thank you for reading The AI MEMO. Feel free to share it with your network.
ONLINE COURSE — Agentic AI: Challenges and Opportunities for Leadership
Prepare for the challenges and opportunities of agentic AI as you lead AI transformation in your organization. Maximize the benefits of these technologies while minimizing disruption.
In addition to the boost in expert-level productivity, the MIT study also found that change (and thereby AI) is making work less enjoyable. More precisely, it is changing how we are used to working. Change is often less enjoyable than maintaining the status quo.
While this aspect of change management is highly relevant at present to increase acceptance and adoption of AI, it won’t be a permanent concern. At some point, AI will be the default way to complete a task, and the current (then old) way of doing things will seem backward.
We need to redefine our understanding and expectations of work, outcomes, and rewards to address it. This wave of AI innovation is not about incremental updates, individual new solutions, or departmental applications. In the long run, this shift will affect everyone.
Yet, we measure expectations and experiences by our current, pre-AI-mass-adoption standards—a trap that applies to managers and employees alike.
It is a key reason why 48% of employees who use AI (instead of completing the task themselves) are concerned about being perceived as less competent in their manager’s eyes. That is why employees feel like they’re cheating when using AI. (See details in Slack’s Workforce Index Fall 2024.)
When technology evolves, our definition of work needs to evolve as well, for example, along three aspects: work, outcomes, and rewards.
Evolving Our Definition of Work, Outcomes, and Rewards
From an early age, we are praised for our accomplishments. In many ways, work—and completing work—is a natural progression for many professionals to feel that accomplishment as an adult. Generative AI threatens that mental model with its rapid pace and broad scope of capabilities.
But here is a proposal for how to evolve it as a leader and individually:
Work: We need to focus on purpose rather than effort. Who are you serving by completing a task or a project? What do we do (as humans), and what do we safely delegate to AI?
Outcomes: We need to emphasize the quality and speed of achieving a result rather than measuring our worth by completing the task ourselves. What are our true strengths? What tasks should we spend our time on in a project?
Rewards: We need to measure an outcome’s impact rather than just rewarding its accomplishment. What has the benefit been? What revenue, savings, etc. have resulted from it?
You can start shaping these aspects today within your teams and individually—you seriously should. Because otherwise, we’ll be judging tomorrow’s capabilities by yesterday’s standards.
Conclusion
AI is helping experts increase productivity. Novices can increase their productivity as well and achieve expert-level results. But one does not simply become an expert overnight. Novices using AI will still need thousands of hours of experience while AI accelerates the learning curve. To maximize the utility of AI, we need to redefine work and our thinking of our own value in the process. Work, outcomes, and rewards are three dimensions that evolve from a subjective (me) to an objective perspective (purpose, quality, impact) and can help us make that leap.
Do your current experts have the skills to succeed in a sea of rapid innovations? Get in touch to learn more in a session or workshop!
Join my bi-weekly live stream and podcast for leaders and hands-on practitioners. Each episode features a different guest who shares their AI journey and actionable insights. Learn from your peers how you can lead artificial intelligence, generative AI & automation in business with confidence.
Join us live
December 03 - Leanne Shelton, Founder & CEO of HumanEdge AI Training, will be on the show to talk about mastering AI in marketing.
