We’re all using AI all the time. We just don’t always recognise it.
There’s a lot of discussion around the introduction of AI-based systems into clinical practice and healthcare systems. But these discussions tend to focus on the systems that are being designed, developed, and deployed as part of formal processes centred on ‘big ideas’ in practice e.g. diagnosis, expert systems, natural language processing, robotics, etc.
There’s another form of AI-based reasoning that doesn’t seem to be on anyone’s radar and that’s the relatively simple matter of search.
Every time a clinician searches for a condition, or a management strategy, or a language translation, they’re getting results returned to them by an algorithm. If they’re signed into the search engine their results will be different to yours. In other words, the ‘answers’ that you get may very well be different to the ‘answers’ that I get, even if we’re using the same search terms. And then we integrate this information into our clinical decision making.
We’re increasingly worried about the effect of AI on clinical practice and patient outcomes, and rightfully so. But we’re only looking at the ‘big’ effects. I haven’t seen many concerns about the millions of small influences being exerted on the thinking and reasoning of clinicians who routinely use search engines as part of their daily practice. While the effects of these searches are probably quite small, their scale and ubiquity make them something worth thinking about nonetheless.
I don’t want to suggest that this is a Bad Thing. If anything, I think that clinicians who regularly look things up are great.
I just think it’s useful to reflect on the idea that, if Google or Microsoft make changes to their search algorithms then the results returned to each of us may be slightly different. And that’s going to influence each of our thinking in slightly different ways.
AI is already a pretty big part of almost everyone’s clinical practice; we just don’t think of it that way. Maybe we should.