Meeting minutes
Collaboration Tools Accessibility User Requirements.
jasonjgw: Janina has put forward a series of revisions.
jasonjgw: I'm in favour of merging into editors draft and editing it from there. Any objections?
jasonjgw: no on
jasonjgw: ...objections heard
jasonjgw: auto summary generation may not be possible for all occasions
jasonjgw: maybe worth adding a note to that extent
janina: the previous note of "rely on human editors" didn't sit well. We shouldn't prevent people using these tools if they may help. We should ask tech to do what it can and trust users to treat the output appropriately.
janina: a note may be more than needed. Just saying "...where possible" or "...where pragmatic" may be enough.
jasonjgw: I'll make the amends accordingly.
janina: it may be helpful to run this past COGA in answer to the concerns they had.
janina: One issue that was raised there was the requirement to be able to rename folders. It's hard to find a platform where that's not the case but there may still be something generic in W3C guidelines that addresses it.
jasonjgw: most of the issues appear to have been addressed.
jasonjgw: we could look at messaging from next week.
janina: there is a messaging directory which is populated.
Artificial intelligence, machine learning, and accessibility.
jasonjgw: this is a recurring topic since it crops up in many of the discussions we have here.
scott_Hollier: this is a common question that we are asked
scott_Hollier: such as "can AI support language simplification?"
scott_Hollier: and "what guidance is there on using AI to support WCAG (eg. alt text and descriptive images)"?
scott_Hollier: we created a doc a while back looking at use cases for different audiences.
scott_Hollier: given we are still getting questions on this we might be in a good place to create a place where these issues can be addressed.
jan:
janina: this is a topic of the day so could be a good area to work on.
janina: 1 how do you control them 2, how do you get a useful output?
scott_Hollier: even the pitfalls of auto captions could be useful to discuss and provide guidance on.
jasonjgw: we do have some relevant guidance.
Raja: often people with disabilities have a lot of privacy concerns and other concerns about this.
raja: for instance auto captions are used by hearing people but often cannot be used by deaf people even those who have good speech.
jasonjgw: there is a certain amount of uncertainty over how these tools will develop.
jasonjgw: maybe an overview of some of the issues would be useful.
scott_Hollier: an approach would be to highlight issues related to the state of the art in terms of using AI to implement WCAG
scott_Hollier: another issue is that technology used to allow non-verbal people to communicate on the phone with an AI voice are being screened out by Banking security software that identifies them as AI based voices.
jasonjgw would a wiki be a good first step?
jasonjgw: all agreed so we'll create that wiki page
jasonjgw: it can be a single page wiki with subheadings for different topics.
Privacy Principles.
jasonjgw: Any updates on this topic?
janina: we'll likely be referring to this from the AI work so it's worth checking this now.
Immersive captions.
jasonjgw: we'll come back to it next week
<janina> DRAFT RESOLUTION: RQTF requests APA take up the Immersive Caption CG Final Report at https://
<janina> logged in email at https://
RESOLUTION: RQTF requests APA take up the Immersive Caption CG Final Report at https://
logged in email at https://
Miscellaneous updates and topics.
jasonjgw: The area of natural language interfaces is shifting due to LLMs it may be worth revisiting that topic.
janina: this could be folded into the AI work.
scott_Hollier: People often don't think about the AI in natural language interfaces because it's not explicitly mentioned.