From Hype to Reality: The AI That Is Really Needed

It will be a real revolution when artificial intelligence changes the core processes of organizations, until then “AI is just a rhetorical exercise”. Alberto Calcagno, who has been on the frontier of the technology for 25 years and now leads a big name in health software like Dedalus, which employs around 8,000 employees across thirty countries, warns against easy illusions about the use of AI in business. “The current hype risks deflating if the head of the AI department continues to be 3-5 steps removed from the CEO: if we continue like that, we are light years away from integrating technology into the true strategy of a group,” underlines the top manager who worked at Fastweb’s control deck for over thirteen years.

Are we wasting time?

Like all technological revolutions, artificial intelligence is driven by technology providers who tell us about how AI will create a different world. But it is the use cases that intercept people's needs that make the difference. If we do not move from theory to practice, we are talking about a rhetorical exercise in technology. In Italy, but I could say that for all nations, eighty percent of the investments attributed to AI go to consultancies, hired by companies to understand how and whether to use artificial intelligence, perhaps in absolutely marginal parts of the business. To ensure that AI has the impact on society that the internet has had, we need to be concrete.

How can we avoid missing the opportunities that the development of information technology makes available to science?

We need to make a leap in awareness and apply artificial intelligence to strategic processes. At Dedalus, there are two macro areas where we are testing ourselves: patient services and internal processes. We are a software company and we develop code: the great challenge especially with generative artificial intelligence is to be able to have AI develop software directly or write code with a human-machine mix to allow our 3,000 developers to increase their capabilities. This means placing the technology at the center of strategic processes and seeking to achieve an epochal impact on the business model: if I can develop at double, triple speed, I can go to market and innovate the entire process much faster for the benefit of our clients, who are the hospitals, and their customers, who are the patients.

If for hospitals the acceleration lies in the software, what does AI bring to the table for patients?

There are two areas of application that we are studying. The first is the one related to algorithms that, based on medical parameters, are able to anticipate the probability of a certain clinical course and manage it in an optimal time with the best treatment for the pathology. A major impact, with artificial intelligence software supporting the doctor who is the one that ultimate. We have certified five different predictive algorithms as Medical Devices and now we will extend it to another 15, and then an additional 30. It’s IT that becomes a form of clinical management, and the software gives you suggestions to intervene medically.

What is the second path you are taking?

We work on diagnostic imaging. AI can facilitate and make the activity of doctors and nurses in treating injuries more effectively, thanks to the algorithm's ability to automatically and three-dimensionally recognize a wound thanks to a simple photograph taken from the cell phone. This allows to save a lot of time (today all observations are recorded manually) as well as ensuring a better quality of care. A solution that is absolutely going in the right direction, because it allows to "give patient time back" to clinicians and at the same time better clinical results for patients.

Why is it difficult to transform established processes, even when innovation offers simpler and more effective solutions?

To do these things you need courage, usually the courage that comes from a CEO. Otherwise, artificial intelligence is made to work on how to optimize the use of office space… In Dedalus, the head of AI reports to me and is also responsible for product strategy. But in this world it can't just be Dedalus doing it, because that's not enough: there must be many more companies that have the courage to work with AI on core processes.

The health care industry has driven many technological revolutions, if you look specifically at your sector, how do you see the future of AI?

Everyone talks about artificial intelligence, but if the question is how many companies have brought it front and center of their strategy, and I would answer that it is less than 10%. The most delicate step, as always, is translating words into actions: you must have the courage to invest, to make organizational decisions. As long as the legal office in companies is more crowded than the AI department, we will be very far away and AI will remain just a stylistic exercise. The challenge, which not everyone will understand, is to overturn the scenario and work to reach the opposite extreme of radical pragmatism. It will not be easy, but if we do not do this, artificial intelligence will not be able to have the positive impact that many expect. Exactly as happened with 5G, for which there was so much anticipation but today is just synonymous with greater bandwidth.