Walk into almost any university lecture hall today and you will find a quiet mismatch. Many students are already using AI tools daily, for explaining concepts, drafting, summarising, translating and coding. Many of their lecturers are still deciding whether to try them at all.
That gap worries me, and not because students are ahead. It worries me because of what it does to the relationship. When a lecturer does not understand the tools, they cannot tell good use from bad use. They cannot guide a student on when AI helps learning and when it replaces it. Rules end up being written from a position of fear instead of familiarity.
The causes are rarely laziness. Lecturers are carrying heavy teaching loads, supervising research and often working with limited time for professional development. Training, when it exists, tends to be a single workshop on a tool that changes before the certificate is printed.
So what would help? First, time. Staff need protected space to experiment, not another task added to the end of the week. Second, practical training built around their own subject, not generic demonstrations. Third, honest conversation. Students and lecturers should be able to sit in a room and say how they actually use these tools without anyone being afraid of punishment.
If we want AI to improve higher education in Kenya, we cannot leave half the classroom behind. Capability has to grow on both sides of the desk.