Don’t Just Train on AI Tools. Help Design Them!

From Task Analysis to Performance Support: L&D Offers the Expertise AI Development Teams Need

AI, AI, AI!

Are you tired of hearing “AI is going to transform the way we work”? For the past few years, the headlines, the coffee shop discussions, and breakroom chatter have all been about AI. Huge fortunes have been invested in the infrastructure behind AI and the tools themselves. So, is it working? Is AI changing the way we work?

Research by The Upwork Research Institute says that nearly half (47%) of employees using AI report that they don’t know how to achieve the productivity gains their employers are expecting, and 77% say that the AI tools have actually decreased their productivity and added to their workload. Why is there such a disconnect between our expectations of AI productivity gains and reality? One key reason may be this: AI tools are being built without learning and development (L&D) having a seat at the table.

77% of employees polled say that AI tools have actually decreased their productivity.

Most organizations, if they think of training at all, assume L&D’s role in AI is to train employees once the tool is finished. That assumption is the problem. When L&D is treated as a downstream training function instead of an upstream design partner, AI tools are built without an understanding of how people actually work, learn, and perform.

The Missing Link — Learning and Development

Let’s look at the tools and how they’re being built and by whom. Overwhelmingly, AI tools are built by development teams that consist of software developers and business stakeholders. Notice who’s missing? L&D, and this absence is a root cause of the productivity disconnect!

If you’re an IT leader, you might wonder why L&D needs to be involved in the process. Usually, when IT builds new products, they bring in the training team at the end to train everybody on the end product. But L&D is uniquely qualified to help build AI performance support tools from the beginning of the project.

Why L&D Should Be at the Table

EPSS

The concept of a tool to support employees performing their work isn’t new. Instructional designers have been studying and implementing Electronic Performance Support Systems (EPSS) since Gloria Gery’s work back in the early 1990s. These systems are computer software programs that improve user performance directly in the workflow. They help people perform work tasks more efficiently and focus on improving performance rather than delivering instruction or information. There are many overlaps between EPSS characteristics and AI tools:

• Just-in-time support and integration: EPSS and AI tools provide help when needed and are embedded in the workflow.
• Context awareness: EPSS and AI tools understand where the user is in the workflow and provide relevant assistance.
• Performance-centered design: They both focus on performance rather than just training.
• Adaptive guidance: They both can adjust their approach or feedback based on the user’s needs and existing skills.

In many ways, AI is the next generation of EPSS. L&D has been preparing for this moment for more than 30 years. We’re ready! Implementing EPSS has taught us how to effectively chunk information, create optimal user interfaces, identify where and when the help is needed, and provide that help in a way that the user understands. These are all things that are needed to create effective AI tools that employees will actually use.

Task Analysis

Instructional designers are experts in task analysis. We know how to break down complex job tasks into manageable steps and identify where help is needed. Many of the negative comments around AI tools are that they don’t solve problems real employees experience. Having an instructional designer observe employees and break down tasks before the AI tool is built ensures that AI tools solve real-world problems and are available to employees in their moment of need.

Learning Retention and Performance Improvement

L&D professionals understand how adults learn and can design experiences that facilitate knowledge retention and skills transfer. When AI tools are created in a way to support workers in their specific moment of need, we can help employees develop their skills in that particular workflow. Instructional designers have proven skills in scaffolding learning. We know how to provide appropriate levels of support that can then be reduced as a user becomes more confident in a skill. Bringing L&D into the design of AI tools will result in a tool that is primed to scaffold information and provide individualized support that can be reduced over time so as not to disrupt the workflow.

Change Management

L&D departments have experience with change management. We know how to introduce new systems and ensure their adoption, and if L&D is involved early enough, we can make sure those systems are designed in a way that facilitates learning for diverse users with varying abilities and experience. Without a proper design background in how adults learn, AI tools fail to achieve desired adoption rates. The cost of poor change management is delayed adoption, shadow workflows, errors, and employee dissatisfaction.

L&D and IT — A Match Made in Heaven

The current approach to AI development that only involves L&D rolling out fully developed AI tools represents a costly missed opportunity. Companies are racing to build tools to help employees perform better and faster, but they’re systematically excluding the very professionals who’ve spent their careers mastering that challenge. Academic research consistently shows that AI tools without proper pedagogical design fail to improve performance.

L&D must have a seat at the table at the very beginning of the analysis and design phase, partnering with IT to build the infrastructure in collaboration with business stakeholders. For L&D, a seat at the table doesn’t mean just helping with rollout decks. It means participating in all phases of the agile software development lifecycle — especially at the beginning where decisions about workflows, prompts, guardrails, and user experience are made. L&D can also help with testing and finally create that training for users.

Call To Action!

If you’re an L&D professional, you should reach out to your IT department and understand what’s in their pipeline. Most IT professionals don’t realize what an ally the L&D team can be before systems are actually built! L&D needs to be recognized as an essential partner, not an afterthought. Our expertise in task analysis, learning retention and performance support, and change management isn’t just nice to have. It’s critical for building AI tools that result in improved performance, not empty promises. This isn’t about expanding L&D’s scope — it’s about reclaiming it.

That’s why ILG can partner with organizations on day one to help integrate your L&D team into AI initiatives, ensuring solutions are designed to truly improve employee performance.

______________________________

Additional Resources

Blog: Embracing AI Tools

Podcast: Revolutionizing Training Using AI?

Webinar: Your Artificial Intern: Using AI to Streamline Design and Development

Webinar: AI: Is IT a Fit in Your Learning and Performance Strategy?