Idea
Have we made our business cases too narrow to recognise some of AI’s most valuable benefits?
Financial returns matter, but conventional business cases can struggle to value wellbeing, judgement, service quality and the avoidance of unnecessary work.
Current thinking
Imagine a clinician saving ten minutes during a demanding day. The benefit might be another appointment, but it might equally be a chance to take a proper break, make a more considered decision or finish work with less pressure. We should be wary of treating only the first of those as valuable.
AI investment decisions often start with hours saved, costs avoided and the prospect of reducing headcount. Those measures are useful, especially when budgets are tight, but they can steer organisations towards the benefits that are easiest to count rather than those that matter most.
Better decisions, staff wellbeing, faster resolution of a difficult case and a less frustrating public service are not necessarily intangible in the sense of being unknowable. They may simply require different evidence, longer observation and a willingness to value outcomes that do not immediately release cash.
If we only fund the benefits we can easily count, we may overlook the changes most worth making.
The open challenge is how to build credible evidence for these benefits without using vague claims as an excuse to avoid scrutiny.
Writing on this idea
Nothing published under this one yet — it is still a question I'm carrying rather than an argument I've made.