What Machines Can (not) or Should (not) Do

In recent years, machines have increasingly taken over tasks that for long only humans were able to perform. Today, algorithmic systems “diagnose” cancer, “decide” who will receive a job or a loan, and provide judges with predictions of who is likely to reoffend. More and more, human judgement is replaced with automated decision-making.

The usage of algorithmic systems can produce important benefits for people’s daily lives. Improved Big Data and AI technologies generate useful insights for researchers, politicians, or companies, and promise efficiency gains. However, the technologies also pose challenges for society and for individuals’ access to opportunities. For instance, algorithmic systems have been found to discriminate against people of color, women, or the poor. Therefore, they have the potential to increase social and economic inequalities.

Delegating decision-making to machines raises fundamental ontological and normative questions: What are the premises and implications of considering machines as agents or decision-makers? Are “decisions” taken by algorithmic systems just and how can they be justified? How to prevent or mitigate risks? And what governance measures are needed to ensure that values such as human dignity or equality are not violated? What role can transparency and openness play?

This public lecture series invites distinguished researchers to talk about the ramifications of new information and communication technologies (ICTs) and how to account for fundamental rights and values in research, design, and deployment of ICTs.

More Information:
https://www.inf.uni-hamburg.de/en/inst/ab/eit/taming-the-machines/winter21-22.html

Wednesdays 18.15-19.45 CET, digital (Zoom webinar) or present (Edmund-Siemers-Allee 1,wing-building west, room 221)