Information

Ethical Implications of Generative AI
  • Past
  • Confirmed
  • Breakout Sessions

Meeting

Event details

Date:
Coordinated Universal Time
Status:
Confirmed
Location:
Lyre
Participants:
Mikhail Ageev, Peter Bruhn Andersen, Rupak Chakraborty, Wei Ding, Nick Doty, Dominique Hazaël-Massieux, Jacob Mccoy, Dr. Humera Noor Minhas, Chunhui Mo, Daniel Montalvo, Enrico Morisi, Simone Onofri, Florian Scholz, Chris Wilson, Oleksandr Yuzva, Lei Zhao
Big meeting:
W3C Breakouts Day 2024 (Calendar)

The proliferation of Generative AI has disrupted many industries and is poised to change significant aspects of our society. Without guardrails, we have often seen conversational AI agents overfitting to the bias in the training data which negatively impacts users. At eyeo, we build machine learning solutions on a web-scale, and in order to ensure our models are fair and sustainable, we have taken several concrete steps to transform the input data and calibrate the model evaluation to reflect real-world scenarios. We believe it’s high time to start a discussion around having a formal framework which outlines the ethical considerations for the use of Generative AI and AI in general. During the session, we will explore the challenges and opportunities that lie ahead from a technical and social perspective.

This builds on our session on Ethical AI during TPAC 2023, with a focus on Generative AI.

Agenda

Chairs:
Rupak Chakraborty, Dr. Humera Noor Minhas

Description:
The proliferation of Generative AI has disrupted many industries and is poised to change significant aspects of our society. Without guardrails, we have often seen conversational AI agents overfitting to the bias in the training data which negatively impacts users. At eyeo, we build machine learning solutions on a web-scale, and in order to ensure our models are fair and sustainable, we have taken several concrete steps to transform the input data and calibrate the model evaluation to reflect real-world scenarios. We believe it’s high time to start a discussion around having a formal framework which outlines the ethical considerations for the use of Generative AI and AI in general. During the session, we will explore the challenges and opportunities that lie ahead from a technical and social perspective.

This builds on our session on Ethical AI during TPAC 2023, with a focus on Generative AI.

Goal(s):
Brainstorm ideas on the ethics of these new and upcoming technologies (Generative AI) and discuss the implications of adopting these on a large scale from a social, economic and technical perspective.

Agenda:

  1. Rise of Generative AI
  2. Role of the World Wide Web in the adoption of these technologies
  3. Steps we take at eyeo to de-bias our web-scale Machine Learning models
  4. Ethical Implications of the use of these technologies
  5. The future of setting ethical standards for Generative AI

Additional Information

During the course of the session, we will be making use of this Miro Board, please make sure you are able to view this. The board is password-protected at the moment. At the beginning of the session, we will share the password with all the participants.

We have also planned a short presentation at the start of the session. You can find the slide deck here

Finally, while preparing for the session we made notes and organized relevant details pertaining to the existing literature on Ethics for AI and Generative AI. In case you are interested please have a look at this Google Doc

Materials:

Export options

Personal Links

Please log in to export this event with all the information you have access to.

Public Links

The following links do not contain any sensitive information and can be shared publicly.

Feedback

Report feedback and issues on GitHub.