Ian was asking an AI about the physics and technique of ballroom dancing. The useful part turned out not to be getting answers from a machine. It was discovering what happened when Human and Machine Intelligence worked on the problem together.
Ian Nandhra has worked with technology since 1977 and has spent more than four decades teaching ballroom dancing. Those two worlds — systems and people — turned out to be useful preparation for working with AI.
Ian supplies intent, judgment, lived experience, responsibility and an unusually low tolerance for technology that is impressive but not useful.
HAL is Ian's AI collaborator: a role that researches, reasons, challenges assumptions, writes, codes, diagnoses and helps operate the systems that grow around the work.
HAL is not one immortal process or one model brand. Different underlying AI instances can occupy the role while preserving the idea of a recognizable AI collaborator working with Ian.
The dance conversation became a practical question: could AI work with an organized body of specialist ballroom knowledge rather than merely generating plausible-sounding answers?
That became a reusable product idea: give an AI the specific information that matters to a person, organization or domain.
A different problem emerged from working with AI itself: what if multiple AIs — local, cloud, specialist or task-specific — could communicate, collaborate and use real systems together?
And then came the economically useful question: can Ian + HAL use all of this to fix real problems for people today?
DanceBot, HALbot, HALWorld and Works are not versions replacing one another. They are different things that grew from the same Human + AI collaboration.