I am slightly sceptical about the great productivity gains suggested for AI. I hope I am wrong but I think it may take quite some time before the ‘intelligence’ of AI gives solid results. That is not to say that AI will not bring benefits, such as productivity improvements, only that we should not expect too much too soon.
There is ,though, one area/industry that seems ready for AI – healthcare.
Healthcare has a lot of bureaucracy and many medical and para-medical staff spend far too much of their time filling in forms, recording data and generally keeping the vast bureaucratic machine fed. If we can reduce this feeding frenzy, we save time and increase job satisfaction (or at less reduce dissatisfaction and frustration).
We have also seen advances in areas such as diagnostic analysis resulting from machine learning (which is what a lot stuff we now call AI really is). Analysing large MRI and other scans is faster and more reliable using AI. AI isn’t perfect at such analysis and subsequent diagnosis but tests suggest its success levels match or exceed those of most doctors.
The next stage is for AI to make recommendations as to further tests or even potential treatments.Such processes will improve over time as AI ‘learns from its mistakes’ and learns from actual outcomes.
AI is unlikely in the near term to take over any ‘hands on’ work of doctors and nurses. But if it can reduce the ‘side load’ of bureaucracy and speed up diagnosis and determination of treatment, we could see quite large productivity gains in the relatively short-term future.