Artificial intelligence can play an important role in the political context

3 Feb 2022

Data-driven methods are very valuable when facing challenges to society, such as climate change.

  • Their use depends on the availability of reliable data
  • The application of AI must be based on clear goals and include procedures for error analysis

Data-driven methods and AI systems can be of considerable use even in a political context. “In complex political situations, artificial intelligence can play an important role. AI systems can support human decision-making, for example, by computing scenarios and forecasts without taking the actual decision away from the humans involved,” said Helmut Küchenhoff, Professor of Statistics in the Faculty of Mathematics, Computer Science and Statistics at LMU Munich, in his “KI Lecture”.

AI and machine learning are already making significant contributions to climate science, explained Küchenhoff. The LMU statistician also sees potential for the use of AI in health research.

Questions around data quality and goal setting

Reliable training data and clearly defined criteria for the goals to be met are needed in order for artificial intelligence to be used, said Küchenhoff. But the LMU statistician warns against having unrealistic expectations, given the particular conditions that prevail in the political context. “We will not be able to solve every last problem with artificial intelligence,” he said, explaining that for many issues, there just aren’t enough data to train the algorithms.

Even having access to data is not enough in itself, he pointed out. In complex situations, the criteria on which decision-making should be based are difficult to pin down, and there are sometimes major uncertainties and constant new developments that need to be taken into account, as we have seen with the coronavirus pandemic. If decisions made with the help of AI are going to be transferrable to the real world, the underlying environment needs to be stable. Nevertheless, especially where the data are complex, algorithms could offer up important results.

The LMU statistician also highlighted the importance of error analysis in AI methods: “It’s a fundamental mistake to believe that AI systems themselves don’t make mistakes.”

The lecture and the subsequent discussion with Prof. Oliver Jahraus, LMU Vice President, and the audience is available on the LMU YouTube channel.

More information and an interview with Professor Helmut Küchenhoff can be found here.

For further dates for the “KI Lectures” and to register, click here:

What are you looking for?