- From: Claudio Tubertini <claudio.tubertini@almalibri.it>
- Date: Thu, 11 Jul 2024 14:24:28 +0000
- To: Laurent Le Meur <laurent@edrlab.org>
- Cc: public-tdmrep@w3.org
======================= Claudio Tubertini Almalibri.it mob +39 327 1503898 https://orcid.org/0000-0001-5768-4499 Thank you Laurent for your continue and valuable effort. Claudio On Thursday, July 11th, 2024 at 4:20 PM, Laurent Le Meur <laurent@edrlab.org> wrote: > I invite you to read an interesting W3C paper from May 2024, titled AI & the Web: Understanding and managing the impact of Machine Learning models on the Web. > > This paper was written by D. Hazael-Massieux and the W3C Team. It "proposes an analysis of the systemic impact of AI systems, and in particular ones based on Machine Learning models, on the Web, and the role that Web standardization may play in managing that impact". As usual, W3C is seeking input from the community and the paper gives access to a discussion repository. > > The section titled Balancing content creators incentives and consumers rights is particularly interesting for content publishers, as it show that their concerns are shared by the W3C Team, and expressed in this document in a properly balanced way. > > The Text and Data Mining Reservation Protocol (TDMRep, initially released in Feb. 2022) is one of the proposed ways offered to publishers for opting-out from crawlers used for AI training (and more generically any TDM processing). > > This paper was the opportunity to discover a similar discussion at IETF, nickname ai-control , with a planned workshop mid-September (2024) in Washington DC. It seems that this discussion followed the release of a post titled Considerations for AI Opt-Out (April 2024) by M.Nottingham. I also discovered the trust.txt specification from the JournalList association, which recently added a `datatrainingallowed` ("yes"/"no") property to their model. > > The TDMRep Community Group will continue working on its specification and exchange with other initiatives, looking for better interoperability between the TDM & AI opt-out solutions that constantly appear in the IT world.
Received on Thursday, 11 July 2024 14:24:38 UTC