When making prediction has some scaring effects

A graphic image used by the Australian newspaper to illustrate the concept of AI at work to predict the life span of a patient. Credit: The Advertiser – Adelaide

AI, Artificial Intelligence, is becoming pervasive. Its progress is no longer based on better algorithms (although they keep getting better), rather on the availability of large sets of data and ways to extract meaning and learn (deep learning technologies).

I just stumble on an article published by The Advertiser, a newspaper in Adelaide, Australia, reporting on a result obtained by a team of researchers at the Adelaide University.

The researchers have applied AI, deep learning, technologies to evaluate the expected life span of patients affected by different pathologies (probability of living for over 5 years).

The AI system was given as set of date the chest imaging (radiographies, TAC) of 48 patients and the system proved to be as accurate as a forecast made by an expert doctor. Interestingly, the way the system analysed the images was not based on a set of rules but on learning from a huge set of data.  Even more interesting, the researchers were not able to pinpoint the specific reasoning that pointed the AI system in giving a certain “prediction” of expected life time.

It seems clear that by observing huge set of data the system can learn to focus on some specific aspects that are relevant for making this kind of prediction. The researchers expect that AI based medical systems will become complementary (not necessarily better, at least in the short term) to expert doctors by looking at different aspects that are difficult to digest by a doctor. The growing availability of digital medical data set will accelerate the learning and will result in better and better systems that will be able to look at broader and broader sets of pathologies.  The important factor here is that the AI system is not just looking at a specific pathology, like a lung cancer, to make its assessment. Rather it takes into account the state of the heart, digestive tract, kidneys and so on, in practise it evaluates the overall “health” of the patient to draw its conclusions.

This is good news indeed. At the same time, the idea of a computer checking on me and spitting out as hard fact my reaming life expectation is a bit scaring. Think about the increasing availability of medical sensors at home and on our body (like the ones in a smart watch) and it is not difficult to foresee that in the next decade as you will be interacting with some sort of distributed computing in your home (through interconnected ambient IoT) the AI available in your home will be checking on you and making some -potentially- nasty prediction.

Of course, it can get worse. What about asking for a life insurance coverage and discover that no one is ready to insure you because their computer has just put up a red light?

Technology, as usual, is nor good nor bad, but it surely opens up unexpected possibilities, some of which we may not like at all.

About Roberto Saracco

Roberto Saracco fell in love with technology and its implications long time ago. His background is in math and computer science. Until April 2017 he led the EIT Digital Italian Node and then was head of the Industrial Doctoral School of EIT Digital up to September 2018. Previously, up to December 2011 he was the Director of the Telecom Italia Future Centre in Venice, looking at the interplay of technology evolution, economics and society. At the turn of the century he led a World Bank-Infodev project to stimulate entrepreneurship in Latin America. He is a senior member of IEEE where he leads the New Initiative Committee and co-chairs the Digital Reality Initiative. He is a member of the IEEE in 2050 Ad Hoc Committee. He teaches a Master course on Technology Forecasting and Market impact at the University of Trento. He has published over 100 papers in journals and magazines and 14 books.