A big part of my work is in following the academia side of things, so I know there is some interesting work being done, but I equally know that the journey from interesting small project to mass adoption is incredibly long. Autonomous vehicles are a good example, as they're pretty good in many situations, but a mile away from roaming London's streets unaided. That final bit is often incredibly difficult from a technical, organisational, legal, and cultural standpoint.
In healthcare, for instance, getting a pilot is relatively straighforward, but deploying something at the kind of scale spoken about to justify the hype is a whole other ballgame. This perhaps explains why startups always talk in terms of the future and the possibilities rather than the reality, but the reality often never comes, so they suffer huge write downs. I mean it's hard to believe that generative AI has caught the likes of IBM by surprise, yet they've effectively mothballed Watson in healthcare because it wasn't delivering any results. You could say the same with the projects undertaken by Google and Apple. Heck, even 20 odd year old technologies like telehealth have largely struggled to take hold, even after the pandemic opened the door to them.