Die neue Infrastruktur des Alltags

What are the central goals of Data Market Austria? What are your experiences so far?

ML: The central goals of the Data Market Austria (DMA) Lighthouse project of the Federal Ministry of Transport, Innovation and Technology are:

• Advancing Technology Foundations
• Creating a Data Innovation Environment

In support of these main goals, DMA is also investigating technology for Interconnecting Clouds and developing Pilots to demonstrate how such an ecosystem would look like.

The two main goals, technology and innovation environment go hand in hand. It is in fact the challenge of this project to make sure that both progress in sync and that they benefit from each other.

We have started our project back in late 2016 and already in early 2017 we had stakeholder workshops across Austria and across industries to provide an initial scope of requirements for such a data market. This conversation with the industry, with start-ups, SMEs, or Large Enterprises has been going on continuously - a results of this is also the collaboration with the Upper Austria Business Agency and its Industrial Data Initiative. The experience so far in this regard is that it is a complex subject, where many have a feeling that it could be something important for them and for their companies, but it is unclear what exactly it is, and what role each can play. I have previously spoken about the value of data and what is clear is that this value is highly subjective. It is a very different market than what most industries are used to. Being digital, it is in some sense similar to the entertainment industry (though few would see this as entertaining). The similarity resides in the fact that once the digital product is out, it is very hard to control its replication and distribution.

This aspect, in the case of potentially sensitive data (and I am not referring here only to personal data, but rather to any data that could potentially reveal internal aspects of a company), significantly reduces the appetite for selling data on an open market in many potential data providers.

Technology can help to some extent here. In DMA we are strongly controlling access to datasets and recording all accesses in an immutable private blockchain, that can be controlled by all providers in a distributed fashion.

However, controlling exactly what happens with the data once the customer has acquired it and transferred it to its own infrastructure is still, in most cases, a matter of research. Methods exist in cryptography, but they are rarely scalable to large data. We are investigating this in the context of another research project, Safe-DEED.

What messages are behind the term digital ecosystem?

ML: In my view the main message is the need for integration and exchange. Digitization is a challenge in itself, hence the need for initiatives like the Digitalization Agency of Fit2Internet but digitalization is not a purpose in itself. It is a means for better life, when talking about the general population, and for better processes, when talking about companies.

To achieve this goal, the digital information we are creating by this digitalisation process must be integrated in an ecosystem, such that, as a whole, society can benefit.

There is a comparison I like to make here, and perhaps it is a bit stretched, but it is pertinent: just like evolution in biology does not happen in isolation, progress in industry only happens if companies interact with each other. The smoother we make this interaction, the more numerous the contacts, the more numerous the adaptations, and therefore the faster the evolution.

The technology adapts to the needs of the user - reality or just a buzzword?

ML: yes, definitely. That does not mean that the user does not adapt to the technology though. That is simply because there is no single user. We are a set of users, again, regardless of whether we are referring to individuals or companies. What is adapted perfectly to me might be slightly less perfect for you, so adaptation is about finding the middle ground – it goes both ways.

At the same time, we need not be naïve about this. Technology developers will adapt to the user because they have a commercial interest in making sure that that user does not go to another provider.

In the case of the data market, our users are companies and we strive a lot to make sure that we develop a technology that is as adaptable as possible to a generic user. Nevertheless, further adaptation and customization will be required.

Is digital in your view the new infrastructure of everyday life?

ML: Again, I have to answer with an enthusiastic “yes, definitely” qualified by the statement that by this I don’t mean that digital will replace all other aspects of everyday life. Also, strong differences are visible between country-side and city-dwellers, between east- and west-Europe, between industrial states and states under development.

For the groups for which I can speak as a member (city-dwellers in a west-European industrial nation) digital is de facto the new infrastructure for everyday life: digital systems control everything from the water quality on our kitchen tap to how we plan our days. Think about this: the basic infrastructure of everyday life is time–how we manage it and how we plan it. Digital time pieces have been around since the 1970s. We have had a long time to adapt to our digital infrastructure.

Is Data Science just a new hype?

ML: As Director of Research Studio Data Science, I have to answer here with an enthusiastic “No”.

But I fully understand the question. We have had, in the history of computer science at least, several cycles where a particular keyword had become so popular that its meaning was confusing or lost. “Artificial Intelligence” is a well-known example, and it is also making its comeback these years. “Big data” was also a recent one.

We cannot deny that the last decade has been proven to be a very data prolific decade. Any quick internet search will generate for you a multitude of plots showing the exponential increase in data available digitally. Therefore, there is a question about the use of this data. “Big data” focused (or still focuses) on our technical ability to process this data. “Artificial Intelligence” focuses on the ultimate goal of processing this data (I will let your readers define what the ultimate goal of AI is). Data Science is about how we get from having and being able to process data to the goals we want to achieve. It is about applying rigorous scientific methods in the process of improving current processes, making them appear more “intelligent”. I think that using the noun “Science” in its name is also a reminder to all data scientists of the fundamental principles that we have to adhere to, including ethical and moral ones. While science in itself does not make moral judgements, scientists are obliged to. Such fundamental principles have been a topic of discussion from the late 1960s when Karl Popper first mentioned them in Vienna, to the present day.

So, while the methods of Data Science (e.g. statistics, parallelization, machine learning) are part of a continuum of research in mathematics and engineering, Data Science as a term reminds us that the great power that comes from having all this information at our fingertips comes with even greater responsibility to use it first for the benefit of society and the protection of the individual.