Attending this event?
The Papers and day two Keynote schedules have not been finalized. As soon as they are, they will be posted on this page. Please check back soon for the full convention program.
Back To Schedule
Friday, October 27 • 1:30pm - 3:00pm
AI ”Townhall” Presented by the AES TC-MLAI

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Recent developments in AI technology (e.g., the most recent release of ChatGPT and other generative AI technologies) are having a profound effect on creative industries. At the time of writing this proposal there is a writer’s strike in Hollywood. Various legal challenges have been launched against companies providing these services. Thought leaders, think tanks, professional organizations and industry associations are calling for action and posting open letters, recommendations, and position papers. Legislators around the world are meeting. As a professional organization representing creative audio professionals, how should AES react?

The AES community is somewhat unique among professional organizations owing to its diversity. It is comprised of audio practitioners, scientific researchers, technology developers and students. As a technical committee (TC-MLAI) and professional organization, what is our responsibility to these different groups? What leadership can and should we as a technical committee and an organization provide?

The AES TC-MLAI proposes holding a “town hall” style event to foster an exchange of ideas among the whole AES community about recent developments in AI and the AES’s role in shaping the future of audio AI. This facilitated, highly interactive discussion will offer insight into the views, opinions, ambitions, and concerns held by our varied members. The event organizers will not advocate for a particular position. Ideas will be captured, and with this information, the AES TC-MLAI plans to draft in late 2023/early 2024 a set of core principles (values), code of conduct, recommendations and/or guidelines (TBD) that can provide leadership on the use and/or development of generative and other AI technologies intended for use in creative audio work. This information may also be shared with other relevant TCs, AES management, etc. who are in positions to address the community’s concerns and interests. Additionally, and more generally, the event provides an opportunity for community members to share thoughts and concerns, which is constructive for community building and building a healthy discourse.

It is important to note the proposed scope: creative audio work. AI is having an impact in many areas of audio, and each deserves due consideration. However, for this town hall, we propose focusing on the use of AI in creative production. The exact structure of the event is to be worked out depending on the time allotted, etc. It is not our intention to control the content or direction of discussions. Interaction with the audience is a priority. However, we also recognize that without structure conversations can quickly become fragmented and even incoherent. Moreover, the issue of AI can be contentious. In planning, we will work towards striking a constructive balance between structure and openness.

One possible configuration might be as follows:

- The scope of the discussion could be limited to 2-3 pre-determined questions about the design and/or application of AI.
- Informative, educational materials will be made available in advance.
- The questions, discussion scope, etc. will be announced in advance of the town hall (in the conference program perhaps).
- Questions may include, for example: “If TC-MLAI were to draft a position document, what kind of document would best serve the community? Should this be a recommended “code of conduct” for audio engineers using AI technology? Should it attempt to guide research and development? Is it information aimed at clients of audio engineers that can help them to get the services they need?
- Other questions might include, for example: what steps, if any, can/should the AES community take to protect jobs and/or the integrity of the work that we do? How can/should we educate novice engineers about the technical and creative potential, the ethics, and limits of AI audio technologies?

To ensure that various relevant perspectives are represented in the room, we will invite certain participants with expertise in, for example, education, standards, commercial product/system development, legal, AES governance, mastering and other stages of the production workflow, etc. We will make it our goal to facilitate a coherent discussion among diverse participants that sheds light on impressions of audio AI now, and also, suggests future considerations and next steps for the AES community.

avatar for Nyssim Lefford

Nyssim Lefford

Associate Professor, Audio Technology, Luleå University of Technology, Associate Professor Audio Technology
avatar for Jonathan Wyner

Jonathan Wyner

Mastering Engineer, Berklee College of Music, Professor of Music Production and Engineering; M Works, Chief Engineer/Consultant
AES President 2021, Jonathan Wyner is a Technologist,  Education Director for iZotope in Cambridge, MA,  Professor at Berklee College of Music in Boston, Chief Engineer at M Works Mastering.  A musician and performer, he’s a Grammy Nominated Producer whose mastered and produced... Read More →
avatar for Gordon Wichern

Gordon Wichern

Principal Research Scientist - Speech and Audio Team, MERL, Sr. Principal Research Scientist
Audio signal processing and machine learning resarcher
avatar for J Keith McElveen

J Keith McElveen

Founder/CTO, Wave Sciences, Research Lead, Spatial Hearing
My technical focus is on solving The Cocktail Party Problem - i.e., separating speech in a reverberant environment with competing (masking) talkers and noise. Applications include audio forensics, hearing assistance, and voice user interfaces.

Friday October 27, 2023 1:30pm - 3:00pm EDT