The AI Skills Dichotomy and the ROI Gap
PLUS: Upcoming events and activities
Check out upcoming events
Dear readers,
Thank you for being here, I appreciate any of the attention you give to these posts.
There are three upcoming opportunities to connect with me:
HR Tech 2024 - I’ll be there next week, please feel free to reach out!
Raising Venture Capital in HR Tech - A private, virtual event put on with ADP Ventures and SemperVirens, taking place on September 30th.
AI in the Future of Work: Agents, debates & discussions, part of SF Tech Week. Taking place October 7th
To be announced event on December 11th, about AI in Reskilling and education
Back to the regular post
The AI Skill dichotomy
Kristen just watches as her attending manipulates the robot’s arm, retracting and dissecting tissue. Unlike the technologies that dot the history of surgery, using the robot makes it iPhone-easy for him to do the whole procedure himself. He knows Kristen needs practice; he wants to give her control. But he also knows she would be slower and make more mistakes, and she’d be going it alone. Slower means more time under anesthesia, which causes strokes. And mistakes mean blood loss, or worse. Hist patient comes first. So, Kristen has no hope of getting anywhere near those nerves during this rotation.
-The Skill Code, Matt Beane
In study after study, the people who get the biggest boost from AI are those with the lowest initial ability— it turns poor performers into good performers. In writing tasks, bad writers become solid. In creativity tests, it boosts the least creative the most… And in a study of early generative AI at a call center, the lowest performing workers became 35 percent more productive, while experience workers gained very little.
— Co-intelligence, by Ethan Mollick
AI, Robotics, and Simulation: Changing the Way We Learn
AI, generative AI, and robotics have been shown time and again to boost the skills of those at the lower end of the experience spectrum. In the context of onboarding new workers, AI can be incredibly effective—helping employees learn basic tasks quickly, creating engaging pathways for learning, and allowing them to be productive from day one.
For example, AI-driven onboarding programs can reduce the time it takes for a new employee to get up to speed, offering step-by-step guidance for routine tasks. Similarly, virtual reality (VR) can train employees on how to use expensive machinery, allowing them to practice in a virtual environment without risking damage to costly equipment. And for industries like aviation, simulation has long been a critical tool—pilots spend countless hours in simulators before ever touching a real plane, mastering the skills they need through repeated practice in a safe environment.
These tools are powerful for initial skill development, but they aren't a silver bullet for ongoing learning. Matt Beane, in The Skill Code, outlines how AI and robotics are breaking traditional apprenticeship learning models that have been in place for centuries. While AI might help someone perform a task more efficiently, it doesn’t push them toward mastery in complex fields like surgery or engineering. This leaves critical skill gaps in more nuanced and experience-driven areas of work.
The Problem with Skills Taxonomy and Skill Mapping
Many companies recognize the importance of understanding the skills and competencies of their workforce. In response, they have turned to skills taxonomies and skill mapping platforms as a way to solve the problem. The idea is simple: if companies can categorize and map out the formal and informal skills their employees have, they can create a clear pathway for reskilling, upskilling, and role placement.
However, while skills taxonomies and skilling platforms are becoming more common, they haven’t yet solved the core issue. Why? Incentives.
Skills mapping might provide a useful framework for identifying gaps in knowledge, but it doesn’t address the incentive misalignment within most companies. Organizations may know what skills their employees need, but without clear, measurable incentives for both the company and the employee, these frameworks fall flat. The investment in learning platforms might tick the box, but it doesn’t lead to the actual development of talent in a meaningful way.
The Investment Gap: Why Companies Aren’t Focused on Upskilling
Despite the proven effectiveness of AI, VR, and simulation for onboarding and basic training, most companies are still underinvesting in upskilling and reskilling their workforce. Employers often prioritize initial skill development—onboarding, compliance training, and operational basics—but fail to adequately prepare their workers for the rapid technological shifts that will demand more advanced skills.
In fact, Ernst & Young estimates that “66% of employment across the US (or 104 million jobs) is highly or moderately exposed to GenAI.”1 While initial skill development is covered, employers aren’t investing enough in the continuous learning necessary to help employees stay relevant as AI continues to transform tasks.
Entrepreneurs and businesses are responding to this skills gap by launching platforms and services aimed at providing skill development. Udemy’s business revenue, for instance, is up 34% year-on-year to over $400M, signaling growing demand. Investors are flocking to skill-based startups that address this emerging need. Yet, inside the vast majority of companies, the process breaks down.
The Disconnect: Incentives and Learning Gaps Within Companies
Despite the growing demand for upskilling, most companies struggle to align their incentives with long-term employee development. While companies may offer learning solutions and categorize employee skills through skills taxonomies, there are few real incentives to encourage ongoing learning.
Most companies would appreciate having a more skilled workforce. They may even offer occasional opportunities to take courses or switch roles when there’s a significant strategic shift. But few companies:
Set aside dedicated time for employees to develop skills
Provide financial rewards, promotions, or clear benefits tied to the acquisition of new skills.
In many cases, companies don't see a direct, measurable return on investment for employee upskilling—unlike consultants, who can increase their fees based on newly acquired skills. Upskilling is often a talking point rather than a strategic priority.
Consultants and Generative AI: A Model for Investment in Upskilling
There are exceptions, particularly in consulting. When consultants learn new skills or earn advanced certifications, the value of their services increases, allowing firms to charge more and capture higher revenue. For consultants, the incentive to upskill is clear and directly tied to their bottom line.
For example, in their Q3 2024 earnings call, Accenture highlighted their investment in learning and development: "As a learning organization and talent creator, we continue to invest in our people, with approximately 13 million training hours this quarter. This averages 19 hours per person, predominantly due to Gen AI, as we prepare our workforce for the infusion of Gen AI across our business in the coming years."
Similarly, companies like Wipro and Tata are making major investments in building AI-ready workforces. By doing so, they can win more deals at higher prices and differentiate themselves from competitors.
Meanwhile, employees at more traditional companies might learn to use generative AI to improve their productivity, but there’s often no direct or measurable payoff to the company. The benefits remain largely intangible or hard to quantify.
The Future: Simulation, VR, and AI as Tools for Upskilling
As AI and robotics continue to reshape the workforce, the need for effective upskilling will grow dramatically. To address the emerging skills gap, companies will need to rethink their learning and development strategies, leveraging the same tools—AI, VR, and simulation—that they use for onboarding, but applying them in more advanced, continuous learning contexts.
Simulations, VR, and AI can help workers practice complex tasks, gain experience with expensive equipment, and troubleshoot scenarios before encountering them in real life. These technologies could bridge the gap between basic task proficiency and deep mastery, ensuring that workers continue to learn and adapt as their industries evolve.
Conclusion: A Call for Prioritized Time for Learning
Many employers offer access to courses (through Udemy or other platform) and mandate compliance training. Employees use of educational courses is rarely reported on, nor discussed as a reason for promotion or a raise.
Compliance training, however, is different. Companies invest in letting the manager, the department leader, and the companies C-suite, who has completed compliance and who has not. Companies should take a similar approach to on-going learning, setting aside time, monitoring it’s use, and provide actual incentives.
In a world where AI is transforming work at a rapid pace, companies can no longer afford to treat skill development as a one-time investment. Ongoing upskilling, enabled by AI, VR, and simulations, is critical for staying competitive. The companies that invest in this future—like consulting firms today—will be better positioned to thrive in an increasingly automated and AI-driven world.
This post was partially inspired by two great books on these topics:
The Skill Code, by Matt Beane
Co-intelligence, by Ethan Mollick
You can also find their substacks at:
The Wild World of Work - Matt Beane
One Useful Thing - Ethan Mollick