2022 Galois Quotidian Machine Ethics Workshop

Much of the high-profile speculation about the future of AI focuses on existential concerns such as trolley problems for self-driving cars, whether “AI overlords” will keep humans as pets, or what happens if a rogue AI decides that all matter in the universe should be converted into paper clips.  

In contrast to the above concerns, we assert that the widespread adoption of intelligent machines into our everyday lives will be driven more by the ability of machines and humans to navigate more mundane ethical concerns such as:

  • Do I have to worry that my AI housekeeping bot is going to gossip about me or share my personal conversations with its other clients? 
  • Can I trust that my AI lawyer is fighting hard enough for my interests given that punishment and retribution are human concepts, and not meaningful for machines?
  • How can I be sure that my AI doctor isn’t in the pockets of Big Pharma?

The Galois Quotidian Machine Ethics (QME) Workshop is designed to explore the foundations of AI Ethics from the perspective of QME. We’ll be bringing together experts from computer science and allied fields to discuss such questions as:

  • Are there protections that need to be put in place before we can widespread societal adoption of intelligent machines in our everyday lives?
  • Are there universal design principles for QME?  Or are these principles domain-specific?
  • How do we negotiate the inevitable tensions between potentially competing goals as privacy versus transparency, consistency of response versus sensitivity to situation?

The ultimate goal of this workshop is to wrestle with the hard (and fun!) questions, uncover hidden assumptions, and to make initial progress towards laying the foundations for “Design for QME”.

Workshop structure:

The workshop is structured as two half-day events, held on Wednesday, February 9th, and Thursday, February 10th from 9:00 AM to 1:00 PM PST (12:00 PM to 4:00 PM EST). Each day will be broken up into:

  • A keynote talk, to kick things off.
  • A series of 10 minute high-impact lightning talks on each participant’s research area and its application to the challenge problem.
  • Hour-long interludes to discuss what we have learned.

We encourage participants to attend both days, but we get it: life is busy.

Outcomes:

The outcome of this workshop will be a joint white paper that summarizes the current state of the field in Quotidian Machine Ethics, including current challenges, the impact of the topic on AI, ML, and Data Science, the current state-of-the-art in the field, as well as highlighting gaps and opportunities in the current body of research. The Galois organizers will edit this white paper and include all workshop participants as co-authors. The paper will be primarily disseminated through Galois’s website.

Attending the workshop:

To register for the workshop, please fill out this short form by January 21st.

Quick note: This is an invite-only event. If you cannot make it but have a recommendation of someone you think would be interested in attending, please let us know by emailing Julie Nelson at julie@galois.com.

About Galois: 

Galois is a computer science R&D company that develops technology to guarantee the trustworthiness of systems where failure is unacceptable. We apply cutting-edge computer science and mathematics to advance the state of the art in software and hardware trustworthiness, including developing methodologies for design, development, testing, and runtime assurance of safety-critical cyber-physical systems. We are an employee-owned company of 130 people, with offices in Portland, OR, Arlington, VA, and Dayton, OH. Since our founding in 1999, we have worked with a variety of clients across government and industry, including DARPA, ONR, AFRL, DHS, and Amazon. To learn more about all our research areas, the 2020 Year in Review provides a great snapshot of what we were up to in 2020. We often organize conferences and workshops similar to the 2021 Galois Summer School for Trustworthy Machine Learning, Artificial Intelligence, and Data Science held virtually on Monday, June 28th, as well as the 2021 Galois Balloween Workshop, which is taking place on October 27th and 28th. We have a very different organizational structure called the Collaborative Web, and you can read/watch more about how we work on Life at Galois.