Philosophy & Computer Science

A Causal Analysis of Harm
Friday, 21.07.2023 at 4:30pm

It has proved notoriously difficult to define harm. Indeed, it has been claimed that the notion of harm is a “Frankensteinian jumble” that should be replaced by other well-behaved notions. On the other hand, harm has become increasingly important as concerns about the potential harms that may be caused by AI systems grow. Indeed, the European Union’s draft AI act mentions “harm” over 25 times and points out that, given its crucial role, it must be defined carefully.

I start by defining a qualitative notion of harm that uses causal models and is based on a well-known definition of actual causality. The key features of the definition are that it is based on contrastive causation and uses a default utility to which the utility of actual outcomes is compared. I show that our definition is able to handle the problematic examples from the literature. I extend the definition to a quantitative notion of harm, first in the case of a single individual, and then for groups of individuals. I show that the “obvious”‘ way of doing this (just taking the expected harm for an individual and then summing the expected harm over all individuals) can lead to counterintuitive or inappropriate answers, and discuss alternatives, drawing on work from the decision-theory literature.

This is joint work with Sander Beckers and Hana Chockler.

Risk Imposition by Artificial Agents: The Moral Proxy Problem
Saturday, 22.07.2023 at 9:30am

Machine Ethics from an AI Perspective
Monday, 17.07.2023 at 4:30pm

In 2006 machine ethics was defined as the field that “is concerned with the behaviour of machines towards human users and other machines”. Since, the terms computational ethics and AI ethics have also been used. Sometimes they are used to denote completely different problems and sometimes different dimensions of the same concern: how should machines behave? From the point of view of philosophy, an oversimplified statement of the machine ethics problem is: what does it mean for a machine to behave ethically. From the point of view of AI what is studied in machine ethics (again putting it overtly simply) is: how do we automate moral reasoning and decision-making? The talk gives an overview of machine ethics from its official inception to today following the work in AI. A related, but pertinent point that also will be discussed is: are AI enhanced machines and softwares de facto acting as our moral arbiters.

Invited speakers are Aldo Faisal, Susanne Gaube, Joseph Halpern, Hendrik Heuer, Eva Lermer, Dimitri Coelho Mollo, Wojciech Samek, Elizabeth Seger, Tobias Seitz, Marija Slavkovik, Johanna Thoma, Mariya Toneva, Kate Vredenburgh, and Herbert Zech.See below for pictures and biographies of our speakers.

Monday, 17.07.2023 – Opening

Tuesday, 18.07.2023 – Intelligent Systems

Wednesday, 19.07.2023  – Fairness & Explainability

Thursday, 20.07.2023 – AI in Healthcare

Friday, 21.07.2023 – Privacy & Security

Saturday, 22.07.2023 – Summary & Closing

Deadline: March 31, 2023

Participation in the summer school is 90€. This includes lunch and coffee as well as a guided city tour.

If you’re interested in joining the summer school, please fill in this document and send it together with your current transcript of records and a CV to phil+cs(at)uni-bayreuth(dot)de. There is a limited number of scholarships (max. 500€ per person) available to support those in need of assistance with travel and accommodation costs. Please indicate in your application if you’d like to be considered for this.

The keynote lectures are public and free to attend for everyone, no registration required. Places available for joing the full program are limited, please apply. There will be no streaming of lectures or keynotes.

Senior researchers who are interested in joining certain lectures or days, please contact us at phil+cs(at)uni-bayreuth(dot)de.

If you are a student at University of Bayreuth, you might be able to take certain parts of the summer school as coursework. For details please refer to CMlife teaching entries; there’s no need to fill in the application form. Of course, you can also apply as a regular participant to join the full program.

The Summer School will include a poster session on Tuesday afternoon. All participants will be invited to present a poster on the topic they are working on.

More information will follow closer to the event.

We’ll help participants to arrange accommodation in affordable hotels. Details will follow closer to the event.

To get to Bayreuth you have three main options:

By plane:

If you plan to come to Bayreuth by plane the nearest airport is in Nuremberg. From there you have to take a train to the main station (Nuremberg airport to Nuremberg main station) and then take the train to Bayreuth (Bayreuth Hauptbahnhof). The route takes about 1h 15min.

Some better connections might also be found to the airport in Munich. From which you can reach Bayreuth by train via Nuremberg. This takes around 3h 30min and you need to change train at Munich main station and Nuremberg main station.

Detailed information about the train can be found here.

By train:

If you will travel by train the best way is to take a train to Bamberg or Nuremberg since Bayreuth, unfortunately, does not have a long-distance train station. From there you can take a local train connecting to Bayreuth. You have to be a little bit cautious to not end up on the wrong train. From both directions, there are two coupled trains from which only one goes to Bayreuth. So, make sure, that you entered the correct train.

More detailed information can be found on the Deutsche Bahn website here. The train station in Bayreuth is called Bayreuth Hauptbahnhof.

By car:

If you come by car, Bayreuth is very easy to reach.  The university is located right next to the Autobahn A9 exit Bayreuth south. There are plenty of parking spots at the university free of charge.