Thread by @WellsLucasSanto

On the last-minute changing of the name: "Rather than say the ways that we would like to deviate from the inevitable, we want to talk about the ways in which the implications of the future are up for grabs." - @alixtrot 🔥🔥

https://twitter.com/FAccTConference/status/1369693836134518786

.@schock tells us to "put our money where our mouth is" and sign up for and support the Turkopticon organizing effort to help support Amazon Mechanical Turk workers:

https://twitter.com/turkopticon/status/1369684074030379011?s=20

.@cori_crider talks about Prop 22 here in CA, which companies like Uber spent $200M on in order to encode into law that drivers are not employees. "Having secured that victory, they're seeking to roll out that model in other legislatures." "That is Uber's vision of the future."

.@cori_crider: "The whole transport economy has been Uberized, and we in the UK have a unique opportunity to bring those companies within some measure of the rule of law. Minimum wage and holiday pay are two concrete consequences. That's one future."

On content moderation, @cori_crider mentions "the rules developed overwhelmingly by white middle-class male Americans are enforced by a completely different class of tech workers" (Look at @ubiquity75's "Behind The Screen" for research on exactly this)

yalebooks.yale.edu/book/978030023…

@VidushiMarda: "One of the most invisible structural challenges that we have as researchers on the ground ... trying to make tech less bad, when I think a lot of the work should be not to have some technologies on the ground at all."

@VidushiMarda: "You're testing & perfecting problematic technology in one side of the world & stalling it for the other kind of the world until it is good enough to be imposed on those people. It creates a deconstructive narrative around what regulation & protection looks like."

@schock channels @YESHICAN ("any intervention that doesn't build political power or agency of the marginalized community is libel to harm rather than the help") and the US disability justice movement ("nothing about us without us") in introducing the design justice framework. YES

@schock: "How are we going to design systems and processes that are constantly working to dismantle the matrix of domination (a term from Black feminist scholar Patricia Hill Collins for the intersecting systems of racism, patriarchy, capitalism, and ableism)."

Btw, @schock just mentioned the Design Justice Network and their principles for "envisioning a future where design is used to support care, healing, liberation, joy, and sustainability." I love the folks at DJN: Sasha, Una, Wesley, Denise, check them out:

designjustice.org

@alixtrot points out a really good thread about the conflict between liberatory thinking and optimization thinking, about how "optimization traps us into certain logics and makes us less imaginative about processes as inclusive as they need to be" Really great insight here.

@cori_crider raises an interesting point: "The risk you run in doing those kind of procedural challenges--let's say accuracy in facial recognition--you legitimize the system and therefore make it easier to survive." -vs- we just say it's inherently problematic & we don't want it

@schock, pushing back slightly: "Gender Shades is such a crucial piece of ammunition for movement organizers and activists who have managed to successfully pass a range of moratoria and bans at state level and potentially federal level"

@schock: "I think it's a mistake to see [optimization and liberation] in binary opposition. How does a particular piece of research move and how does it get used by broader movements, whether those movements are explicitly abolitionist or not?"

@schock, on the values of redefining optimization questions for our 'technical' community: "What are we optimizing for? We could be optimizing for strong consent in AI and machine learning. That has many components that are technically challenging."

@schock mentions there's a whole converesation about what it would mean to translate questions of consent to the technical space, and points at the Consentful Tech Project (which I've actually been thinking *a lot* about in my work). Highly recommend:

consentfultech.io

@cori_crider mentions context in the public sector, and how government "hides contestable policy choices behind a technical veneer", pointing at the UK grading algorithm that was a huge political mess.

theverge.com/2020/8/17/2137…

@cori_crider: "It's on all of us to say, uh-huh, it doesn't have to be this way. To go to Sasha's point, you optimized for what? If you ask people in the A-level, well, we optimized for maintaining the curve instead of individual fairness. People would have said, no, thank you."

@VidushiMarda discusses exclusion that happens with pushback: by the time academia or civic government is able to push back, it's already late in the process and it's often after these technologies are deployed by these corporations. And this happens internationally.

@VidushiMarda: "When it comes to pushback, it's almost like you're given a losing hand. Everything is already decided and done. You can send in comments in 30 days, but you don't know whether they're being read." 👀👀👀

@VidushiMarda: "I think thinking about power more critically, not just in terms of how technologies are used and who they're used on, but how we talk about technologies, where we focus our energy on technologies would be excellent."

@schock: "We need to shift the conversation from fairness to justice and equity. We have to shift the conversation from bias to harm. I encourage everyone to check out the film "Coded Bias" that helps do that."

codedbias.com