"Digitalization is a social learning process"

Ms. Hechensteiner, which medium do you prefer to learn about new things in the world of science?

Sigrid Hechensteiner: Right now, podcasts, because between work and family I hardly have time to read. I can also listen while folding laundry, walking the dog, on the way to work. The other day I was so engrossed in the podcast "Science Versus" that I left the house without my eyeglasses - and I am really blind. (laughs)

And what about social media?

Hechensteiner: I use them in communication, both professionally and privately, but with great caution. I know that content is suggested to me by algorithms. And I certainly don't want to move around in filter bubbles that prevent any form of constructive dialog with people who think differently.

Mr. Laner, information is now more than ever determined by technologies. How dangerous is it when IT experts direct the communication flows of millions of people?

Dietmar Laner: That's dangerous, of course. It's not for nothing that social media is increasingly caught in the crossfire of criticism. In social media’s early days around the Arab Spring in 2011, it was celebrated by social scientists as a tool for mobilizing against autocracy and dictatorship; today, authoritarian regimes are taking advantage of these platforms as multipliers of fake news and abstruse narratives; Western democracies are being undermined. This is also because social media is programmed by private companies who are driven by profit.

Hechensteiner: In retrospect, governments believe that companies like Facebook and co. should have been more tightly regulated. Now they want to make up for it, but what the regulation should exactly look like is the subject of heated debate. The fact is that social media law is very young and has always lagged somewhat behind IT developments. And whether fake news can be legally prevented - I'm not sure! Freedom of expression is fundamentally the right to opinion, not the right to proven knowledge.

How can knowledge be proven?

Laner: Through transparent disclosure of data and analysis procedures. When researchers publish a result today, they increasingly also provide the data sets on which they based their result; and they disclose the methods they used to evaluate the data. This makes it more important that the data sets are collected and digitized cleanly.

This is where your team comes in, Mr. Laner?

Laner: Together with the researchers, we IT experts are looking for the best possible information technology solutions to organize, store and visualize big data. In the "International Network for Long-Term Ecological Research in the Matscher Valley-Valle Mazia" project, for example, we built a platform that displays all data, such as temperature, humidity, precipitation, and wind for the various measuring stations, almost in real time.

And once the knowledge is proven, then it is up to the communication to find an understandable narrative.

Hechensteiner: Which - as it turns out in times of pandemic and war - is extremely difficult. We are in the midst of a battle of narratives, which is primarily being fought out digitally: the web is flooded with false news, or, in extreme cases, people are denied access to information altogether. Today, science communication is challenged more than ever. On the one hand, disinformation must be prevented, and on the other, dialog with dissenters must be maintained.

What role will artificial intelligence play in data collection and processing in the future of science?

Laner: Now, the hype around AI is greater than its possibilities. In the future, however, it will be possible to leave process steps in science to machines. The question we will then have to ask ourselves is primarily an ethical one. Which decisions or work processes do we want to leave to software and, which do we want to continue to make or execute ourselves? Eurac Research is already using AI in administrative processes. This includes the ticketing system, software that is used to request internal services. An AI ensures that every request ends up with the right contact person.

Hechensteiner: AI will also play an increasingly important role in communications. In 2020, the first AI speech production system came onto the market. It’s called GPT-3 and can compose and supplement texts itself with minimal input. The self-learning system can reproduce human thought and reasoning patterns so accurately that the texts generated by the machine can hardly be distinguished from those created by humans. A wonderful tool for all those who must frequently write similar texts, I'm thinking of meteorologists; but a worrying tool when used by online trolls. An article on the subject has just been published in "Nature", in which the author warns of the loss of trust that research will suffer if AI speech production is not regulated.

What are the advantages of digitization?

Laner: It takes a lot of pressure off research management. We are now able to digitize complex administrative processes. This gives us valuable time that we can invest in our core business - research - and it creates transparency. Thanks to digitization, research has picked up speed enormously.

Hechensteiner: We can now reach far more people in our communications. Via the Eurac Research website, our socials and blogs, we can share unfiltered research results first-hand and enter into dialog with society. The number of webinars we put out has tripled in the last year alone. In addition, interest in research is high in South Tyrol, especially among the young public. We used to be perceived as an ivory tower, yet digitization has now opened many doors.