W3C home > Mailing lists > Public > public-silver@w3.org > July 2019

Re: TED talk on algorithm bias

From: Alastair Campbell <acampbell@nomensa.com>
Date: Thu, 11 Jul 2019 16:37:54 +0000
To: John Foliot <john.foliot@deque.com>
CC: Silver Task Force <public-silver@w3.org>
Message-ID: <A3FF02A1-614B-43DA-B633-F9F36F743168@nomensa.com>
Hi John,

> Rather than ignoring that bias, we need to be aware of it, and then hopefully act accordingly.

Agree, any structure (like conformance in WCAG 2.x) and any structured calculation (like a spreadsheet of scoring) will prioritise certain factors.

The key is being aware, and making the calculation explicit and understandable.

You seem to have taken against the WoMD video, or the application of it’s points, but I suspect if you read the book you’d probably agree with a lot of the conclusions. E.g. update the ADA so companies can’t use health or reputation scores a stand-in for medical exams (which are prohibited). The book’s focus is around ML & hidden algorithms rather than what we’re doing.

Personally, I’m not sure what the fuss is about here.

Anything we produce must be open, reviewed, and if we do a good job it will be understandable and explicit about the decision making process that went into it.

I think Léonie’s email that just arrived shows the sort of open thought process that we should engage in to arrive at a solution.

Is anyone arguing against that? Or is this a strenuous agreement?

Cheers,

-Alastair
Received on Thursday, 11 July 2019 16:38:19 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:31:46 UTC