Today, two tweets have rushing through my timeline. Both touch on the subject of digital diagnostics and they show an idiosyncratic, seemingly contradictory picture. Does digital diagnosis lead to better results – or a diagnosis by a human doctor? And from whom do we want to receive such a diagnosis?
Digital diagnosis as a life saver?
The Intelligent Health AI from Basel is bringing positive news, enthusiastic about the feasibility.
One could object: Where is the news? The fact that the diagnostic capability of halfway modern systems of artificial intelligence is superior to human expert knowledge should come as no surprise. This imbalance is well documented. Every oncologist, radiologist and probably almost every laboratory doctor will confirm this. Even if the ideas of the consequences probably differ considerably: The fact is indisputable.
Analog students at MIT?
Futurist Andrew McAfee paints a different but remarkable picture of his practice at the university:
Contrary to all reason, it seems that the next digital elite – nothing else is being trained here at MIT – is actually putting up with disadvantages. They opt for human diagnostics and not for digital diagnostics. He does not comment on the motives. Even if he did, this would hardly lead to a statistically reliable picture.
Three models of interpretation
Let us place the two impressions next to each other and interpret them together. Three patterns of interpretation seem plausible:
Interpretation 1: In case of doubt, technological fascination is always the solution for the others. Autonomous driving is as inspiring as it perfectly makes sense, if only one’s own steering wheel does remain. I call this the deficit model of technological disruption. The guiding principle is the fear of losing familiar solutions, services and features, despite all the technological fascination.
Interpretation 2: The time lag shows an apparent contradiction. This is the model of harmonization over time. Today, students reject what they will get used to over the coming years. At first glance, this is an obvious idea. At second glance deceptive: Those who allow themselves to be guided by this interpretation are in danger of covering up the disruptive character of innovation with harmony sauce.
Interpretation 3: In this juxtaposition we clearly see how one of the most important innovations in health care will take place. The model of the creation of meaning through innovation. A conventional diagnosis that does not involve life and death will – with good reason – be made and communicated by a human doctor in the foreseeable future. Even if an AI would actually be better, there is no real risk. However, when it comes to the threat of fatal diseases, AI offers a leap in quality; those who want to overcome this threat are less choosy when it comes to choosing the means. The main thing is: something works, even if it´s digital diagnostics.
Disruption does not begin with harmlessness
If this third approach prevails, we will see AI in use very soon. The triumph of digital diagnostics, however, will not begin in the harmless and risk-free, but – on the contrary – where it really counts: In the fight against life-threatening diseases. Ebola, malaria, rapid cancer will bring AI into the everyday life of healthcare before it also devotes itself to the fight against colds and lice infestation. Sometime later.
I tend, you will suspect, to the third interpretation. The most important innovation of healthcare of the future will begin with questions of life and death. But please, judge for yourself.