Luca Gambardella, the IDSIA is the research institute for artificial intelligence in Lugano. What should we understand by this; what projects are you currently working on?
The institute was founded in 1988 following an initiative by the Italian philanthropist Angelo Dalle Molle. In 1999, it was affiliated to the universities of Ticino. We act as a bridge between academic and applied research; on the one hand, we work on research projects supported by the Swiss National Science Foundation, and on the other on application-oriented projects with the Commission for Technology and Innovation (CTI). We also carry out many activities with the European Union and direct mandates with companies around the world.
Let us dare to take a look into the future. What possibilities will artificial intelligence open up?
Artificial intelligence or AI set the market in motion a few years ago; the biggest global players are all heading in this direction, from Google to Facebook and Amazon, IBM, Tesla, etc. as well as banks. We research the implementation of systems that have ever greater intelligent capacities. The market demands this, and in my view the potential is great. It is already possible to develop machines that can do something better than a human being - image classification or the detection of production errors, for example. However, these machines still need to get smarter. A machine that is more intelligent than a human being does not yet exist.
How has AI changed the economy and what changes are still to come?
The economy needs quick solutions; it wants machines that help experts with their work. Human-machine interactions are widespread, but even if intelligent machines help people to make decisions, I do not foresee humans rapidly being replaced by machines, at least not in the area of AI. It’s more likely that this replacement will occur for mechanical activities, since these are a matter of traditional robotics. However, AI primarily deals with machines that autonomously make decisions, are able to learn and configure themselves; the digitization of the economy will require a lot of artificial intelligence.
Which sectors of the economy are particularly interested in your work?
Certainly the financial sector for the areas of economic data analysis, stock trading or the creation of profitable portfolios. There is also the worlds of media, fashion, online shopping, tourism or leisure. In addition, we try to optimize machines with production companies and work with doctors to understand medical data. Language analysis is another important area: this is concerned with automatic translation and everything to do with chatbots - text based dialogue systems - or vocal assistance, in other words the communication between man and machine.
Many economic sectors are interested in AI. Which one is making the biggest progress in your eyes?
I believe that the fastest advances can be achieved in application areas in which it is possible to use big data applications to automatically train artificial intelligence, such as self-driving cars, risk analysis, online support, insurance, marketing, in short everything that uses big data in some way. A lot of progress has been made by companies that develop their own systems; these include drones, which then become robots that can make their own decisions.
Could you give us a few specific examples of how Swiss SMEs use AI?
At the moment, we have ongoing projects with about 15 SMEs. The Swiss Federal Commission for Technology supports cooperation between universities and companies in order to bring innovation into companies. There are many interesting examples: it's not just product manufacturers, transformation companies and the financial and medical industries that approach us, but also hotels, tour operators and newspaper publishers. Everyone tries to understand how data can be better utilized when it comes to interacting with people. The important thing here is to make decisions. It's not about experimenting with AI, but putting into practice where the business is.
What does IDSIA do to remain a top research center?
We invest in human capital, in order to remain our attractiveness as an employer of highly-skilled people, and also in machines. Complex machines are essential for certain AI projects. The education of students and postgraduates is an important cornerstone. We see ourselves as researchers who develop new ideas and translate them into applied projects. I am convinced that research should never stop studying and inventing new things. But a practical disposition is also needed to apply this in the real world.
More information on the topic can be found in our dossier.