Meeting minutes
notes agendum 1 added
notes agendum 2 added
jasonjgw: Notes email discussions re references, some re restructuring
jasonjgw: Believe we're here to document anything further
sees 2 items remaining on the agenda:
1. Accessibility of Machine Learning and Generative AI. [from jasonjgw]
2. Miscellaneous topics. [from jasonjgw]
shawn: Would like to sync up on where we are and were we want to be
shawn: What our next steps are
shawn: Want to get in sync on goals and approach
shawn: Points to email from july 2024
xi> present+
Kushalnagar (~RajaKushalnagar@a7a68d30.publics.cloak) has joined the channel
shawn: is goal to reach out and get more contributions?
jasonjgw: I can add what I think consensus currently is
josh: Can also comment on that
josh: have been working on how to grapple with a vast topic
ey (~stacey@a7a68d30.publics.cloak) has joined the channel
josh: We have heard scope creep is a huge problem
josh: everyone has an opinion on what we should address
josh: We wanted wider conversation to help define what scope would best help the community
jasonjgw: agrees with that summary, believe we have agreement in tf
jasonjgw: describing as issues paper was apt
jasonjgw: iterating issues was helpful -- rather than providing guidance
jasonjgw: Sec 2 Scope is currently our understanding
jasonjgw: Discussion is ongoing
RajaKushalnagar: I like the categories so far; it makes sense; understand there can be more opportunities
RajaKushalnagar: to comment on different needs
RajaKushalnagar: we started that
sees no one on the speaker queue
Janina: notes that the open issues are legitimate. In defining what is needed by way of maturity to satisfy criteria for publication of a draft isn't clear. W3C characterizes a first public working draft and a final Note, but no others.
sees shawn on the speaker queue
Janina: notes that first public working draft tends to solicit comments, but intermediate drafts less so.
janina: Concerned to better understand what constitues sufficient maturity for fprocess identified status of updated draft--only two, fpwd and final
josh: Appreciate Jason's characterization of this document as an issues overview
josh: also agree with Janina on not getting caught up in scope redefinitions
sees shawn on the speaker queue
jasonjgw: Some inevitable in this area because the tech is changing quickly
jasonjgw: Notes josh's proposal opens space for some of what we have yet to address
sees shawn on the speaker queue
sees shawn, Kevin on the speaker queue
shawn: Happy to help drive contributions to the work
shawn: Wanted us to announce scope and announce -- as soon as comfortable, not wait for FPWD
shawn: Confident we'll get the contributions wanted if we do that
shawn: Still need to understand purpose of this publication and especially of how it fits in other work
shawn: How is this different; How does it add in the broader community
shawn, you wanted to respond to Janina
sees Kevin on the speaker queue
sees Kevin, shawn on the speaker queue
josh: Sounds good; but moving fast and we're also trying to figure that all out as we go
josh: These documents are useful as they are;
josh: What are the issues? What are the gotchyas?
sees Kevin, shawn on the speaker queue
josh: We're trying to catchup the tech; and to inform long standing a11y veterans the implications
josh: we should perhaps figure out what this isn't
josh: there are ethical/existential concerns we need to address
josh: we need to address quality;
josh: in output, in tools
kevin: will we ever catch up? and if never, what's the purpose of this doc? how it recognize that and position itself?
kevin: don't disagree, but don't see in the doc
kevin: clarifying this in the doc would help
kevin: currently difficult to see that it's there
sees Kevin, shawn on the speaker queue
sees shawn on the speaker queue
josh: Hoping the Editor's Draft; as opposed to the static for the CfC
josh: believe clearer in the ED
shawn, you wanted to repsond to Josh and to respond to Josh on the pace of AI and to say not a note?
sees no one on the speaker queue
shawn: definitely the pace is a huge thing
shawn: is a Note about AI feasible given pace of change?
sees janina on the speaker queue
shawn: We don't have to go Note track; we have other ways of sharing accessibility information from W3C
sees no one on the speaker queue
Janina: the issue of fast-moving technology has arisen previously and will recur. Janina suggests the human needs are not going to change, and we are clarifying that the technology may change but the needs of populations won't. Defining what that viewpoint leads to (e.g., are the results of AI plausible) is what we are currently doing. Some areas we haven't adequately defined, e.g., personalization opportunities.
Janina: it will take time to include these issues. Doing so will be a useful product of the group.
janina: Notes tech may change but humans not
sees no one on the speaker queue
Kevin: asks josh if comments in response to issues raised?
Josh: no, we've been discussing
Josh: it's been there for a bit
Kevin: if we're starting the conversation in ways not yet been explored?
Kevin: are we then giving anything actionable?
sees no one on the speaker queue
janina: Sorry about my caps lock! unintentional!
jasonjgw: enduring concerns and opportunities --
janina, you typed too many words without commas; I suspect you forgot to start with 'to ...'
sees janina on the speaker queue
sees janina, shawn on the speaker queue
jasonjgw: we want to give a better understanding to people developing specs and best practices
sees janina, shawn on the speaker queue
Janina: agrees with enduring concerns and opportunities.
Janina: the title is known to be inadequate. We chose a "less is more" strategy originally.
janina, you wanted to say I like that as a title
sees shawn on the speaker queue
jasjosh: agree it could be like that
josh: agree we can get there; pub right now is an invitation for scoping conversation
josh: quality is perhpas not there enough, but we'll get there on all of this
sees shawn on the speaker queue
shawn, you wanted to ask start and motivation and to comment AI Accessibility available from Canada
sees no one on the speaker queue
shawn: not fully understanding motivations for the doc and how it's different from what's already being done in the community
for example ASC-6.2 Accessible and Equitable Artificial Intelligence Systems
Accessible and Equitable Artificial Intelligence Systems - Accessibility Standards Canada
jasonjgw: Much bias conversation in the literature, but not about effect on pwd
sees no one on the speaker queue
sees no one on the speaker queue
Janina: acknowledges that bias has been discussed but suggests more work is needed,
Shawn: bias and disability is already being addressed [e.g., search web for "ai bias disability")]
Kevin: +1
josh: tooling another interest -- another aspect
josh: I don't see it being covered
shawn: notes industry marketing "we'll help you meet wcag3"
shawn: agree misinforamtion and disinformation. maybe note not best way to address that?
jasonjgw: could be lost in the conversation is to dev conceptual framework can be applied
sees shawn, Kevin on the speaker queue
sees Kevin on the speaker queue
jasonjgw: guidance on how to think about
jasonjgw: a framework to help people think more clearly about how
kevin: believe there's no difference of what we've been doing for years
kevin: can we write what do we need to do different to accomodate
kevin: concerned we're not making community support clear
josh: that's what we're trying to work out
josh: if there's any group out there in a position to comment on what ai is doing right and wrong, it's this one
shawn: want to make sure we're coordinating and not stepping on toes
josh: notes this is different from our customary aur pubs
Shawn: I think you'll be pleasantly surprised by what W3C has already published
AI & the Web: Understanding and managing the impact of Machine Learning models on the Web , August 2024 , W3C Team Report https://
also from W3C : Artificial Intelligence (AI) and Accessibility Research Symposium https://