Synchronization Accessibility User Requirements.
jasonjgw: Timed Text has been pinged; still looking for a response
jasonjgw: One editorial note is left about a summary
janina: Need to confirm if we need a formal CfC for a second wide review
jasonjgw: keep the email exchanges about announcements going
jasonjgw: We should do our own review at this point as well
jasonjgw: We can come back to this next week for a resolution for publishing
Accessibility of Remote Meetings.
jasonjgw: Reviewed document and made some editorial changes in a branch
jasonjgw: Tried to clarify the text and made a few editorial corrections
jasonjgw: Will merge those changes if no objections
janina: No objections to the merge
janina: What about timing?
jasonjgw: Give some time for review and publish after the SAUR
janina: How long of a review period for this and the SAUR - we need to decide on that
jasonjgw: Probably at least 4 weeks and maybe 6, but let's decide next week
Review of draft ethical machine learning consultation document.
jasonjgw: Request was made for comments on the draft
jasonjgw: Refers to UNESCO position and makes an extension to W3C
jasonjgw: A meeting is planned for discussion but date has not been set
janina: We need to be sure to pull in Jutta Treviranus on this discussion
jasonjgw: Even defining what fairness means in the context of machine learning is complex and disability has not been widely considered
jasonjgw: There are other types of biases which have received more attention in the literature, but disability bias is a complex area
janina: Our goal should be to find some way to influence the development of AI so that disability bias does not get introduced
jasonjgw: The primary guidance so far is to include people with disability in every step of the development process
jasonjgw: Should note that even if bias is not initially ingrained in the development, machine learning can become biased as it is deployed and "learns" biased information
jasonjgw: We have not seen much regulatory interest in this so far
jasonjgw: Bias is of course problematic in human judgement as well; some propose that machine learning could be employed to limit such biases
Natural Language Interface Accessibility User Requirements.
jasonjgw: Continuing work on open issues
jasonjgw: Josh merged the changes he was working on last week
jasonjgw: Let's get back to this next week
Inaccessibility of CAPTCHA.
janina: A meeting is planned with CloudFare later today
jasonjgw: Work of the Fido Corporation as discussed in a paper (will share with list)
jasonjgw: Proposes cryptographic keys that follow the user on any device they access
jasonjgw: The maturity model work is being organized
jasonjgw: As mentioned on the mailing list, the APA GitHub repository has been reorganized
janina: The COGA work plan is up for CfC
janina: Daylight Savings Time for Europe comes into play next week