How powerful are algorithms

ITA dossiers to read more

Internet technologies store and record our interests and preferences. But what if algorithms and automated structures have a say in whether we should get a job offer or a housing contract? Computer models also influence political decision-making. Blockchain technologies promise to revolutionize the way we receive and process information.

Algorithms have become a strong driving force in society. The Austrian Labour Market Service (AMS) is currently testing an algorithm to help finance training measures. Based on statistical analysis of previous years it estimates probabilities of finding a job in the future. The aim is to target job seekers for whom the support measures are most likely to lead to a job. But which criteria are used to determine who belongs to which group? What if algorithms serve to discriminate on the basis of gender, health conditions or place of residence?

Similar questions apply to search engines. Google results are anything but neutral: gender and skin colour play just as important a role as marital status and buying behaviour. For example, it has been proven that men will be offered better job advertisements than women. Such content distortions are consequences of a complex interplay between Big Data, algorithms, economic interests and social practices. This needs to be examined more closely. The ITA carries out projects that critically question the social ordering power of algorithms, computer models and search engines.

Europe in the hands of the Big Four

Many of the popular information technologies are operated by US based tech giants such as Google, Amazon, Facebook and Apple ("GAFA"). We also ask how Europe can position itself in this race: How should European policy regulate global technology companies whose business models are primarily based on "data trading"? How can alternative internet technologies and infrastructures be created that are oriented towards the common good and are based on European values such as privacy by design? And finally: Do we need new political bodies and institutions that address socio-political and ethical aspects of algorithms and Big Data applications, and what role can the EU play here? 

Policy advice through computer modelling

Computer modelling can provide insight on urgent social issues by considering a multitude of factors: How can we master the impending energy transition in the most cost-efficient way and with the lowest CO2 emissions? How will a possible influenza epidemic spread and how can it be stopped? Big Data applications allow for more data to be fed into existing computer modelling than ever before, with the purpose to create a more accurate forecast of the future. In debates on scientific policy advice, however, the social influence and might of computer models is rarely analyzed.

We investigate to what extent certain social structures, ideas and values are reinforced with the modelling process. For example, is it really advisable to only look at the best economical solutions when Europe models trade policy with Africa? Are these results an objective, realistic representation of the world, or do certain values go unquestioned? Furthermore, it is essential for policy advice how this kind of knowledge is translated. Do computer models give politicians the message that they only have to take the 'optimal' decision calculated, or do they offer enough information to decide on different options? The ITA asks what significance these models have in everyday political decision-making.

The blockchain revolution

Blockchain technologies promise transparency, security and trust. Blockchain algorithms enable the creation of network-based databases. Information is stored indefinitely and is always accessible to all members of the same network. An example: Blockchain technology was first widely used for the Bitcoin cryptocurrency in 2008. Information on each transaction (e.g. exchange of crypto money) is automatically recorded and managed. But application possibilities are almost unlimited, as every industry has a need to document its transactions securely. It promises so many new opportunities - for international trade, government and finance, among others - that some people almost refer to it as the "next Internet".

However, if these potentials are to be realised, challenges remain. It is important for any new, possibly even life-changing technology to avoid a situation in which a few private companies take absolute control of the data and the market in the end, thus, leading to a covert monopolisation. From our experience with GAFA we know that it is very difficult to control big data companies. The blockchain therefore requires a thorough technology assessment: Which values guide the innovation processes of blockchain companies? How do they differ from public values? And how can we ensure that blockchain innovations contribute to the common good?

Selected projects related to The Politics of Information Technologies