NSF Convergence Accelerator Workshop for Ethical Design of AIs (EDAIs)
Please fill out as many questions as you like. Name is optional.  Please try to give a detailed answer to one or two questions.
Sign in to Google to save your progress. Learn more
Name

Discuss a system or application where some ethical flaw became apparent, either during testing, or after deployment. (A citation or URL is helpful)

What was the ethical issue and why was it not anticipated or addressed originally? With some a priori knowledge, could this failure have been prevented? 

What are some “known unknowns” (i.e., open challenges) in order to move the needle on the ethical design of AI?

What are some ethical issues that remain the hardest to address? 

Provide some examples. What are the prospects for making progress on these issues?  Which gaps should be prioritized in terms of new solution approaches?

 

Describe an ethical design principle / technique / methodology that is known in the research Lab but is not commonly deployed in applications. 

Why has it not been deployed? Conversely, are there ethical practices that are deployed in the wild but that are not known or well-studied in the Lab?

What are the different “levers” that could be applied to incorporate ethical principles into AIs?

How will you know that you have succeeded in incorporating ethical principles into your AI? 

What approaches are the most effective and helpful in shaping and improving ethical design? 

Is the current distribution of responsibilities across different stakeholders (AI scientists; HCI scientists; data scientists; coders; product managers; ethicists.) in the ethical design of AIs appropriate and effective?

What changes are needed? What groups need to play a different role and what would that be?
What other question should we have asked and (optionally) what is your answer?
Submit
Clear form
Never submit passwords through Google Forms.
This content is neither created nor endorsed by Google. Report Abuse - Terms of Service - Privacy Policy