Preparing the Workforce for AI? Start Here

In 1974, the National Council of Teachers of Mathematics (NCTM) boldly stated that all grades should adopt calculators.

As calculators shrank and got cheaper, NCTM confidently predicted that, with a calculator in hand, students would rely less on memorization and “enlarge the scope of problem-solving.” Lofty goals for a device that runs on a watch battery.

The calculator’s mainstream adoption raised questions about how teachers would add the device to their curricula and how students would use it to improve their math skills.

Generative AI’s (GenAI’s) mainstream adoption raises questions for business leaders and employees. McKinsey researchers surveying global companies found that 72% have adopted artificial intelligence in at least one business function.

A 2023 analysis by the World Economic Forum found that GenAI will impact jobs in two main ways: automating routine tasks and augmenting tasks that require abstract reasoning and problem-solving skills. These impacts present significant challenges for leaders who must balance how to use AI to automate tasks while developing employees’ AI skills to work alongside it.

According to a survey from recruiting software provider iCIMS, upskilling employees is a need that “keeps CHROs up at night.” You can hear in those worries the echoes of mathematics teachers having to reconfigure their lesson plans to accommodate the new learning styles that calculators would make possible. Just as those teachers adapted, so will organizations need to help employees adapt by providing training and development to give workers the eclectic mix of skills they need now that AI has increased “the scope of problem-solving.”. To implement effective upskilling and reskilling programs for the AI era, business leaders must focus on three critical components: skill mapping, skill measurement, and skill development through AI integration. Let’s explore these components in detail, followed by actionable do’s and don’ts.

Mapping Skills to Jobs

The first component of an effective upskilling and AI integration program is mapping skills to existing and emerging job roles using a skill taxonomy.

A skill taxonomy provides labels and definitions showing how skills group together. Andrew McAfee, principal research scientist at the MIT Sloan School of Management, classifies GenAI as “a general-purpose technology.” Consequently, employees need a diverse set of skills, such as “adaptability,” “critical thinking,” and “scientific research,” to effectively use and collaborate with AI.

Before requiring or developing these skills in employees, organizations need company-relevant definitions that HR departments can use to map skills to job roles. Although this process requires time-consuming job analysis and ongoing maintenance, it is crucial for effective upskilling. Thankfully, HR leaders can get started with pre-built models and software tools, including GenAI, which can save time and effort on maintaining their models and skill mappings with outside consulting help.

Far from being administrative busywork, skill taxonomies are the essential foundation for accurate job-to-skill mapping. Relevant mappings based on clear definitions provide the “what,” identifying which skills are relevant for each job role, and the “why,” explaining how each skill contributes to job performance.

Measuring Skills

Once an organization identifies relevant and necessary skills through mapping, it needs to measure their prevalence among its workforce. Effective skill measurement inventories employees’ skills to identify where gaps exist; once employees have been trained, it visualizes how an organization’s training investments have impacted employee development and return on investment.

Still, rapid skill changes are forcing managers and trainers to move beyond traditional measures, such as experiences and knowledge-based assessments.

While self-assessments and peer reports can kick-start skill measurements, the best skill verifications come from hands-on demonstrations, such as simulations, structured interviews, and work sample assessments. Each of these assessment types can be difficult to develop, validate, and maintain.

The capabilities of AI-powered assessments offer new ways to overcome these challenges, but buyers should beware. Personnel assessment experts Sara Gutierrez and Richard Landers recommend that leaders should “not trust [assessment] vendor claims without additional research, in large part because vendors in this space often have much greater expertise in AI development than in HR practices.”

Developing AI Skills

Once an organization has identified critical skill sets and has methods to measure them, there’s one more component of an effective AI upskilling and integration plan: developing AI skills with hands-on experience.

On the one hand, employees have already started to adopt GenAI tools. Research and consulting firm Gartner estimates that over the next four years, “75% of enterprise software engineers will use AI code assistants, up from less than 10% in early 2023.” On the other hand, employee-led experience comes with its own challenges, such as data security and privacy.

Some companies have found that IT-approved tools, such as Microsoft Copilot, can be difficult for employees to navigate and tie to real business outcomes. One chief technology officer recently made headlines for canceling its company’s six-figure access to Copilot because it “didn’t provide enough value.”

To successfully adopt AI tools and develop requisite skills, HR and learning and development (L&D) leaders must properly identify tasks well-aligned with the technology’s capabilities. Daniel Russo, a computer scientist studying GenAI’s impact on software development, put it this way: “the compatibility of AI tools within existing development workflows predominantly drives their adoption.” In other words, there needs to be a strong fit between the task and the technology.

Once a company selects the right tools for important tasks, its employees need to be given clear expectations through policy, guidance, and support.

Practical Do’s and Don’ts for HR and L&D Leaders

AI’s impact on work is broad and enduring, defying easy solutions. But there are some practical do’s and don’ts that workplace leaders can follow to overcome the challenges inherent to upskilling employees to work alongside AI technology.

Do Start with a Task, Not a Job

Rather than focusing on which jobs can use AI tools (or which teams need to develop AI skills), identify critical tasks that work well with AI. As Sam Altman, founder of OpenAI, once quipped, “AI is good at tasks, not jobs.”

Integrating AI at a task level not only helps prevent employees from seeing AI as a threat to their jobs, but it also overcomes the challenges of knowing which skills to target for development and which skills can be augmented with AI tools. This task-centric approach increases the likelihood of success for both human workers and machines.

Do Have Good AI Demonstrations

Demonstrations, case studies, examples, and hands-on workshops all help demystify new technology and show how humans best work alongside AI.

Ask those in your organization who adopted AI early on to facilitate these workshops and demos. When employees see colleagues working with AI successfully, they’ll more easily accept the new technology and internalize the lessons they learn during training.

Do Encourage Failure (Within Policy Boundaries)

Starting small allows HR to iteratively design a clear AI policy that works while also providing employees with the flexibility to adapt to rapid technological changes and emerging best practices.

With clear policies for AI use in place, leaders can encourage safe experimentation with AI tools on a larger scale. This allows employees to coach each other, share success stories, and get the hands-on experience they need to increase their confidence.

Don’t Rush to Purchase Learning Content

Without first understanding the most pertinent skills and their benchmarks, teams can waste thousands of dollars on training platforms and materials that develop the wrong skills.

Instead, identify the new set of skills a team or function needs for an objective and the AI tool’s use case. Equipped with these skills, companies can purchase or produce learning content that is best aligned to target tasks and L&D objectives.

Based on the wording, it’s unclear whether they need to -identify- the AI tool’s use case, or if the team needs to -learn certain skills based on- the AI tool’s use case. Can you clarify?

Don’t Start with the Entire Organization

Many leaders quickly get overwhelmed when mapping, measuring, and developing skills for every job in the organization.

To avoid overwhelm, focus on a single job function, department, or team. Take the parts of the process that work to a larger team or identify ways to streamline those parts. This makes the subsequent steps more efficient and generates positive word of mouth among employees.

Don’t Forget About the Culture

Let’s say your team figured out the technical details, skills, technology, and learning content your company needs; it could still all go wrong if the company culture is ignored.

Pluralsight’s Developer Success Lab, in studying AI’s adoption among software engineers, discovered that companies with L&D cultures were much more likely to successfully adopt and work alongside AI.

Any technology might feel threatening to employees if leaders fail to take stock of the culture and whether it’s competitive or developmental (even on a small scale). If the culture is competitive, employees will likely see AI tools as a threat, snuffing out a successful adoption before it even gets started.

Summary

While not all teachers in the scenario we examined earlier added calculators to their classrooms, by 1997, 95% of students were using a calculator on the Scholastic Aptitude Test (SAT). Since then, standardized test scores in math have increased, and researchers have found positive benefits for schools that have integrated the calculator into their curricula.

Companies that can integrate GenAI into their workforce by mapping, measuring, and developing new skill sets will reap the rewards. But, unlike the calculator—a device specialized for mathematical computation—GenAI is a general-purpose technology that needs more intentional steps from users to successfully integrate it with work.

These steps include, 1) starting with individual tasks at a team or departmental level to create demonstrations and learn lessons that companies can use with larger functions, 2) waiting to purchase or develop training content until workplace leaders successfully identify critical skills and suitable tasks to automate or augment, and 3) considering the current state of the company culture. With a culture that encourages some trial and error and peer-to-peer learning, adapting to the AI era will be simple math.