AI in Housing, Welfare & Social Care Network

On how bias and stereotypes were first baked into artificial intelligence: nytimes.com/2019/09/20/art… ⁦@AI_HoWS⁩ ⁦@ATDFourthWorld⁩

Choices previously made by humans are increasingly being made using algorithms and this risks exacerbating existing inequalities. @rachelstatham_ explains...

.@RMaskos Does posthumanism take into account power hierarchies when we empower non-human structures such as #artificialintelligence ? @AI_HoWS @iHumanSheff @EducationSheff #ihumansymposium2019

DJL's Joanna Redden has contributed to this new report for the Centre for Data Ethics and Innovation on online targeting where she details some of the data harms we need to address.

We could not have wished for a more engaged, knowledgeable, and enthusiastic group of people sharing knowledge at our inaugural event today! What a great experience! Thank you all who attended and contributed! @SheffSocScience @iHumanSheff #AIHOWS

A true pleasure meeting everyone &hearing different perspectives @AI_HoWS workshop today #ArtificialIntelligence in #PublicSector @JennyJHayes @SheffSocScience @EducationSheff @iHumanSheff

“AI systems are often designed to predict risk, but what if they could be designed to predict and then support hope?” #AIHOWS

Time for part two of our discussions: what do we already know and what do we need to know? What should our research priorities be? @stephenpotter A real privilege to be joined by people from different sectors and with #LivedExperience to think about these questions. #AIHOWS

Lots of discussion about AI seems to assume that the people on the receiving end of services actually want the services they are being identified as needing, and that they trust the decision makers in making them. #AIHOWS

When we’re talking about data we’re talking about peoples’ lives, sometimes about very difficult things in their lives. @ATDFourthWorld #AIHOWS

The quality of the data depends on the people recording it - these people already aren’t able to collect all of the important correct information for families so how can their recorded data be any better? Can you imagine what AI is going to do? @ATDFourthWorld

Discussion time for @AI_HoWS @SheffSocScience workshop - how do we feel about AI? Can do lots of things in name of efficiency but for whom is it efficient & is efficiency the bar by which we shld judge/measure? Impt to have ppl with lived experience involved in development

“Many of the concerns about the risks of AI are risks that are already prevalent in the systems we’re talking about.” @MattLariv #AIHOWS

How can people be involved in meaningful ways in the development and continued assessment of the use of AI in public services?

“Is efficiency the bar in which we should be assessing AI? Is it a supply-driven market? Are we just implementing AI because we have the technology - not because we need it?” @FieldsDesiree taking talking points between academics, people with lived experience, and local govt

We’re back and in break out sessions to discuss how people feel about #AI, and how its use and output can be critically appraised! Then what #research priorities do we need to respond to? #AIHOWS @digitalsocsheff @SheffSocScience

Transparency of AI is a key issue in 2 ways, explain @JennyJHayes & @MsSarahGibson 1. Transparency of decisions & 2. Transparency in adoption of technology @AI_HoWS @SheffSocScience