AI between comfort and loss of control

Data-driven decision-making models are regarded as important pillars of digital business. But are algorithms that make decisions a purely technical matter? We talked with Stephanie Fischer, speaker at solutions.hamburg, about the impact of such models on company culture.

mgm live: Steffi, what was it that led you to deal with data-driven decision-making models?

Steffi: My focus was originally on change management and organizational development. And one of the biggest drivers of change in companies in the context of digitization is currently the use of big data technologies. In many of my customer projects, automated decision-making processes were explored and introduced. At the same time, the question arose: What does this really mean for the organization and the people within this organization?

mgm live: How do businesses deal with the new opportunities that Big Data technologies offer?

Steffi: Companies are aware of the fact that employees are not strictly rational beings and often make erroneous decisions. At the same time, globalization is exerting pressure with entirely new competitors. This is where the intelligent combination of data and AI comes into play. It allows companies to make better, faster decisions – be it fully automated or as a support for a decision that continues to be made by people.

Many companies are currently taking the first steps – but without considering the consequences for their operation to the full extent. They understand the algorithm only as a tool to become more efficient, faster, better as a company. They use their resources primarily to improve this tool.

mgm live: What are the most frequently ignored aspects? And why is that problematic?

Steffi: People need to learn to deal with these new tools and need new skills. For the forestry workers, the introduction of chainsaws was an immense work relief, but in the wrong hands it can cause great damage. Likewise, companies must decide who can use which tools.

This development is also controversial as a whole. Who decides what an algorithm may decide? For example, should artificial intelligence be involved when deciding whether Hartz-IV is followed up or whether the unconditional basic income comes into force? And what dimension does the algorithm optimize – debt reduction, full employment or happiness?

mgm live: What can be done to consider the opportunities and risks more precisely?

Steffi: It is important first of all that the delegation – that is, the issuing – of decisions must not take place to zero or one hundred percent. For each use case, there are many degrees in between. Do I want to give the entire control to an algorithm or only a part? A wider discussion must also take in companies with representatives from different sectors and hierarchical levels and in society with representatives of as many interest groups as possible.

Big Data and AI are techniques that solve not only technical problems. They influence the lives of people all over the world. Machine learning is already taking away many unpleasant routine tasks. On the other hand, it also makes decisions for us, for example in the classification of Fake News or in the personalized selection of search results.

The central questions are: What is still comfort – and what is already loss of control? How much control have we already given to artificial intelligence? How far does automation need to interfere with life?

Affected parties – regardless of whether they are a company or a whole country – need appropriate skills to understand and assess these technologies. Where and how these technologies are used should not only be in the hands of a few specialists.

Share