Since early 2023, we have been integrating AI into our programmes – beginning with prompt-based exercises and progressing to fully customised AI assistants. Late last year, a client asked us to design an AI assistant for a customised programme – but with a catch: they wanted it to be based on their internal AI platform. Three of our senior trainers worked with their IT department to accomplish that. Then came a surprise.
Our trainers introduced and fine-tuned the AI assistant over several runs in the first quarter of this year. Initially, the results were mixed. After the AI assistant’s issues were resolved, our trainers noticed something else: participants became overly dependent on the tool, overlooked gaps in their reasoning, and felt overwhelmed by the volume of AI-generated content.
So, we redesigned the experience: human-led learning first, AI support later.
This simple shift made a remarkable difference. Once participants had fully grasped the core concepts, it enabled them to use the AI assistant to test, refine, and elevate their arguments – resulting in sharper thinking, clearer outcomes, and a boost in confidence.
The takeaway is clear: when it comes to embedding AI into learning experiences, timing is just as important as the technology itself.