Episode 14: Why HR Can't Sit Out the AI Conversation

AI tends to be guarded by the technology function, treated as the natural property of CTOs and engineers. Holly Joint and Ewan MacLeod think that is a mistake, and this episode makes the case that the people who shape how AI affects our working lives shouldn't be the last ones in the room.The starting point is a talk Ewan gave with their colleague Mike, an HR specialist, to a roomful of senior chief HR officers in the Gulf, mostly from financial services. Mike's concern was blunt: HR is not steering the introduction of AI: IT is. And when the people function isn't leading, it risks becoming a passenger while every consequential decision gets made by the AI team and the chiefs. The room itself proved the point. Asked to rate their own AI fluency, most people placed themselves at two or three out of ten, and rated their organisations lower still.Rather than lecture, Ewan demonstrated. He showed the group Claude's agentic tools taking control of a desktop, opening a browser, researching, and building spreadsheets and presentations from a typed prompt. Then he set a single one-line instruction running: build an OKR system for a bank in the UAE. The pair moved on to other discussion, and roughly twenty minutes later a complete, working system appeared, seeded with its own demo data, polished enough that the HR experts recognised it instantly as something they would normally pay a great deal of money for. That recognition, Ewan notes, was the moment the experts came alive, suddenly seeing the engineering, procurement, training and people dimensions all at once.Holly broadens the argument beyond HR. If these tools are going to reshape work, they can't be built and governed by engineers alone; doctors, lawyers, accountants, artists and people specialists all need a voice in how they develop. Her warning to HR is direct: get knowledge and hands-on experience first, because without it the function will be left behind and seen as irrelevant. You cannot have a credible view on how work will change if you have never used the tools doing the changing.The episode's most practical thread is where to begin. Holly favours auditing first, mapping individual capability and, carefully and anonymously, the shadow AI already in use, then honestly assessing organisational readiness rather than leaping to "give me a use case." Both hosts champion experiential learning over theoretical presentations: let people play, have fun, and lose the fear. And crucially, don't skip the top. Boards and executives are too often handed compliance training when they carry the greatest accountability for AI-driven decisions, especially in regulated industries. Start there, then the exec team, then everyone else.Key TopicsWhy AI shouldn't be owned solely by the technology functionHR's risk of becoming a passenger in AI adoptionThe case for a broader range of voices in shaping AI toolsA live demo: a bank OKR system built from one prompt in 20 minutesEmployees outpacing their organisations in AI capabilityAuditing capability, shadow AI and readiness before adoptionExperiential learning over theoretical AI presentationsWhy board and executive training should come first