July AI/ML Community Group Call: Respectful reuse and AI/ML Tags in IIIF Manifests

47 views
Skip to first unread message

Emma Stanford

unread,
Jun 26, 2024, 3:14:43 PM6/26/24
to IIIF Discuss
Hello everyone,

Our July 24 AI/ML Community Group call will be themed around how we can advocate for respectful reuse of IIIF resources in an AI/ML context. Allison Sherrick from METRO.org will be reprising her wonderful presentation  with Diego Pino from IIIF 2024, and talking in more detail about their proposal for including "no-AI" or "regulated AI" tags in IIIF manifests. Then we'll have time for discussion of these topics. We hope to see many of you there.

Call information:
July 24, 10am San Francisco | 1pm New York | 18:00 CEST | 17:00 BST
https://stanford.zoom.us/j/92701885540?pwd=U0JxQUFGUm1hLytTMk9tYzBWUC9GQT09
Meeting ID: 927 0188 5540
Password: 219585
Agenda and notes: https://docs.google.com/document/d/1N69Q6eW2faXZYUMh0XVdMoEFF_SOHb7HNDeouWpSWr8/edit 

Abstract for Allison and Diego's IIIF 2024 presentation:
We are living through yet another transformative era when the fundamental drivers of Internet content creation, curation, and reuse are shifting as massive AI and ML applications reform our shared digital landscape. In response to these shifting sands, we need to rethink the ways we can best provide quality research and reuse interactions for our richly described content shared through IIIF APIs and related Manifests. The unregulated consumption of data and resources by AI-powered bots and harvesters is increasingly more tenuous and challenges our well established ideas of what openness means. We propose the implementation of standardized “no-AI'' or “regulated AI” tags in IIIF Manifests that could be applied in repository environments across the globe. Similar to the initiatives crafted by other creative content communities to limit the scraping of images and text using “noai” and “noimageai” meta HTML tags, we propose that the IIIF Community considers implementing a set of tags within the standard IIIF API frameworks to help better regulate AI/ML content scraping and non-consented or attributed use of IIIF powered content in AI/ML applications. We strongly believe in the value of creating consensus and shared practices in our own community for combating this significant challenge.

Allison's bio:
Allison Sherrick, Digital Projects and Services Manager
Allison works as part of METRO’s Digital Services Team to extend METRO’s digitization, metadata, hosting and support services for digital collections and custom digital projects development. She is an active member of the Archipelago Commons open source software team, and specializes in implementing and supporting diverse digital repository configurations and environments. She is enthusiastic about supporting and expanding access to information resources from institutions of all shapes, sizes, and geographic locations.

Stephen White

unread,
Jul 11, 2024, 6:58:47 AM7/11/24
to iiif-d...@googlegroups.com
What would be the difference between "regulated AI" and "curated AI"?  Knowing that the current state of the art would require some curation of AI output for use as reference material.


--
-- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
---
You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/iiif-discuss/4ca1d5be-d36f-4484-a95d-950bc5635a4an%40googlegroups.com.

Emma Stanford

unread,
Jul 11, 2024, 11:20:20 AM7/11/24
to iiif-d...@googlegroups.com

Hi Stephen,

 

I believe that “regulated” in the context of Allison’s work refers to the regulation of AI input, not output—that is to say, limits set on how much and for what purposes the material can be used by AI. But please attend the call in two weeks’ time!

 

Thanks,

 

Emma

 

Emma Stanford

Digital Services Manager

Hoover Institution Library & Archives

Stanford University

650-723-1558 / Tower 221

she/her

 

Reply all
Reply to author
Forward
0 new messages