Really appreciate opportunity @TransportforNSW partnership has provided for #utsmdsi students to tackle real-world challenges in such a welcoming collaborative context
Debriefing from my session with our @UTS_mdsi#DataScience students on #sensemaking. Not everything can be quantified. The power of different modalities. Tools for deeper dialogue. Visual Explorer from @CCLdotORG https://t.co/PBFjvoL1u9
Flags the challenge of creating socially responsible Algorithms @baxterkb: Algos & AI rely on the data they're ‘fed’. If the data is inaccurate, incomplete, unrepresentative, regardless of how “good” the AI, outcomes can perpetuate & expand inequities #RightsTechAu https://t.co/chM22et1CR
Really appreciated your #RightsTechAu presentation yesterday @manishaamin about the need to look at the edges which recognises the value of 'outliers'. Echo your point that having choice is not the same as having power. Look forward to further conversations and collaborations. https://t.co/oovGgNoeJB
Yesterday's #RightsTechAu conference provided a tasting plate of the challenges of our data-rich digital futures. Today at @AusHumanRights workshops we will begin a deep dive into ways we can raise ethical AI. Excited to be part of these important conversations.
Argh. As if to illustrate the point that our digital, data worlds rely on energy, I'm about to run our of power on my laptop. So I shall lose my twitter voice because I didn't bring my power source :-) to #RightsTechAu https://t.co/i982DH8YcR
Thanks @feraldata at #RightsTechAu for bringing up the challenge of dependencies and desires for data-rich, energy-hungry systems and implications for our future. There is a growing community exploring sustainable AI https://t.co/dg7yuT1nPn
Thanks @feraldata at #RightsTechAu for bringing up the challenge of dependencies and desires for data-rich, energy-hungry systems and implications for our future. There is a growing community exploring sustainable AI https://t.co/dg7yuT1nPn
Would love to get @solomonbrett to visit our campus and share more of these examples with #utsmdsi students as they learn to practice human centred #DataScience#TechRightsAu https://t.co/HH5l6kOuhH
yes. Great way forward. In #utsmdsi we ensure students experience what it's like to be represented in data & outcomes of machine-informed analysis. Remaining alert to ways data represents human experience sensitises coders & analysts to the power of their judgments #RightsTechAU https://t.co/HH5l6kOuhH
Panel 2 #RightsTechAu sees a conversation about innovation now looking at inequities of the digital divide. If our future is digital & data informed (a question in itself) and we wish to empower citizens with choice, we need to better educate people to truly empower them
what I'm hearing in this second #RightsTechAU panel is that in a civil society we should be putting effort into building informed, data literate citizens -- informed "data publics". Citizens, rather than consumers
Building on @baxterkb observation in her earlier #RightsTechAu talk -- imagine what would happen if all the talent currently focussed on AI innovation that feeds consumerism were focussed on critical issues like child protection and domestic family violence. https://t.co/wNEzylKQ48
I think algorithy will become a new favourite of mine. Let's use it in a sentence: "On reflection, the autonomous car's latest decision seemed particularly algorithy" Lite relief in the midst of serious talk about innovating in the interest of human value at #RightsTechAu https://t.co/WKTS80NSN8
My two favourite word creations from the real-time #AI speech transcription system on the #rightstechau main screen so far 1) enkrimtion and 2) algorithy
excellent point raised by @TobyWalsh#RightsTechAu about need for more ethical review of tech practices. Worth remembering as Hannah Wallach flags in her wonderful Medium article: big data is really granular social data https://t.co/sLHi0Sww8Z
Perhaps the lesson from this morning's #RightsTechAu panel is that we humans should think twice about outsourcing all our decisions to machines. Let's become better partners with AI options. Let's make sure humans are learning as our machines our learning so we can "build better"
bias isn't always visible. Some of most biased judgments can be made when we think we're not being biased. Multiple voices (& machine/human partnerships?) would however have to be be better than a single voice/perspective. It's why we triangulate in social science. #RightsTechAu https://t.co/Z4xTyNK75W
Ultimately technology is just a tool. Humans set the course. Humans create the biases that they then pass on to their technological creations. #RightsTechAu