Artificial intelligence is poised to fundamentally transform our economy. Thirty percent of hours worked today could be automated by AI by 2030, with efficiency gains spanning nearly every industry and role. Some estimates suggest AI will change the way 8 in 10 US workers do their jobs. Business leaders are rushing to take advantage of this technology, with the vast majority of CEOs either making or planning significant investments in generative AI.
But there’s a catch: before they can harness the power of AI, companies need to upskill their workforces. Right now, almost two thirds of workers say they lack the skills to use generative AI effectively, and about the same share of companies say their biggest skills shortages are in AI and machine learning. Equipping employees to build and use AI tools should be an urgent priority for every business. Companies that do this well will have a major competitive advantage. Those that do it poorly will risk creating lackluster AI products and using AI tools in ways that compromise ethics, equity, and privacy. Those that don’t do it at all will be left behind.
At Uplimit, we’re already working with tens of thousands of learners—at companies ranging from Fortune 500s to innovative growth-stage startups—to build high-impact AI upskilling programs. And although it’s early yet, we have a good sense of what it looks like to “do it well.” Here are four principles we believe will be essential to long-term success.
Create dedicated learning tracks for AI end users, builders, and experts.
When designing an AI upskilling program, it’s helpful to segment employees into three groups. Investing in training that meets the distinct needs of each group will maximize the benefits of AI across the organization.
The first group is non-technical end users, including marketers, designers, HR managers, etc., who can use AI to improve the efficiency and quality of their work. Training for end users should focus on building comfort and familiarity with AI tools. Many non-technical employees are intimidated by AI or believe it’s not “for them,” which means simply demystifying the technology and demonstrating its practical applications can have a profound impact on productivity.
The second group is AI “builders”—technical employees, such as software engineers and data scientists, who are or will be building AI-powered products. Thanks to the emergence of flexible open-source AI models over the past year, any employee with a programming background can learn to develop AI products in a matter of months, dramatically increasing the return on AI upskilling for this group. For companies that plan to integrate AI into their products or build internal AI-powered tools, just six to eight weeks of training for most technical employees can be a game-changer.
The last group, AI experts, includes advanced technical employees who are actively developing and honing AI models. Training for this group can be more fluid, with the goal of keeping these employees up to date on advances in cutting-edge AI research and applications. And finally, employees across all three groups should complete training in AI ethics, safety, and company-specific policies.
Prioritize hands-on learning.
The goal of any upskilling program is for employees to apply what they learn on the job, so it makes sense that the most effective programs would include practice solving real-world problems. But when it comes to AI, hands-on learning isn’t just a best practice. It’s essential, and it should form the backbone of program content.
This is largely because AI is such a foundational and versatile tool—Bill Gates has described it as “as fundamental as…the personal computer, the Internet, and the mobile phone.” Think about how you learned to use the Internet; it almost certainly wasn’t by listening to lectures. Instead, you were given an Internet connection and tasks to complete, from sending a message to researching a topic to mapping a route to the store. By working on those tasks, you developed intuition for what the Internet was capable of and how to use it to achieve your goals. The same principle applies to AI.
Hands-on learning is also critical because AI is so new. There’s no set playbook for using this technology, and every application is ripe for optimization. Giving employees space to experiment and find their own solutions is one of the best ways to move the field forward. Creating space for employees to interact and learn from each other is key, too—seeing how colleagues solve problems with AI can spark ideas that boost efficiency across levels and functions.
Implement fast and iterate constantly.
AI technology is changing almost daily, which means upskilling programs have to evolve just as fast. A program that takes months to design will be obsolete before it’s implemented. Instead of focusing on polish and optimization, learning and development leaders should focus on flexibility and speed—rolling out good (but not perfect) content, keeping it current, and iterating as they go. The goal should never be to create evergreen resources, but rather to ensure there’s always a next step for employees to take.
At Uplimit, we update our courses every time they run, which usually means once every few months. Courses in fast-changing fields like Deep Learning have been completely overhauled multiple times since they launched. Creating new courses is a constant priority, too; recent additions to our catalog, such as Fine Tuning Large Language Models and AI Strategy for Leaders, teach in-demand skills that were niche or nonexistent less than a year ago. To facilitate rapid iteration, we deliver most of our content in easy-to-update formats (text as opposed to video) and continually invest in user-friendly course authoring tools.
Iterating quickly also requires continually sourcing and acting on feedback. For L&D professionals this means staying in close communication with functional team leaders, who will be the first to know whether training content is working for their teams and when new AI developments have implications for their work. AI itself can be a powerful tool for gathering and distilling large volumes of feedback; for example, Uplimit uses generative AI to collect insights from learners in each run of a course and extract takeaways to improve the next iteration.
Use AI to teach AI.
Education is one of the many industries AI is poised to transform. AI models excel at some of the most time-consuming tasks in teaching, including creating first-draft instructional content, adapting content for different audiences, delivering feedback on assessments, and identifying learners who need extra support. Programs that lean on AI for these tasks will be able to evolve faster and serve a wider range of employees—in short, they’ll be better able to implement the first three principles above. And the content and learning experience in these programs can serve as another example for students of how AI is applied in the real world.
When evaluating AI-enabled training options, L&D leaders should look for programs that use AI to enhance human instructors, not replace them. AI is a powerful teaching tool in large part because it allows expert teachers to spend more time on tasks that require a human touch, such as motivating and inspiring students. Programs that leverage the best of both AI and human instruction will be far more effective than those that rely solely on one or the other.
AI upskilling is both an unprecedented challenge and an enormous opportunity for learning and development leaders. Every employee needs to understand the basics of how this technology works and how to use it. But every employee and function will engage with it differently, and best practices will change continuously as the technology evolves. Designing a program that uplevels an entire organization, and keeps everyone’s skills current, will require a new level of strategic focus. Leaders who do this well won’t just improve productivity for individual employees and teams—they will drive positive transformation across all areas of their business.