Our 2017 paper Critique and Contribute: A Practice-Based Framework for Improving Critical Data Studies and Data Science is now one of the most cited in Big Data. Open Access here: liebertpub.com/doi/full/10.10…
With Anissa Tanweer @BrittaFiore@lilaavati
We show how people working with data grapple with pragmatic ethical choices in their day-to-day work. We do this by looking at Data Science for Social Good teams, academic data science teams, and energy engineering teams.
We look at four common critiques of big data
data are inherently interpretive
data are inextricable from context
data are mediated through the tools that produce them
data become a chance to negotiate values.
We see these versions of these critiques voiced by the data scientists and model makers whom we study.
The implication is we can't just teach data scientists an ethics course and be done with it. That's not what the data science teams that we studied need.
What they need is help with the process of identifying and clarifying the best practices of making data and doing data science as they reflect on their work. And they need help in designing new kinds of ethical institutions to support this kind of work.
At least among the values-driven teams that we studied these data science and engineering teams already want to be ethical. Their problem was figuring how to make it so.
We call for two practical things going forward to, um, make it so: 1) Create new opportunities for data science practitioners and data science scholars to engage in sensemaking together
(2) Support new kinds of organizational arrangements to foster a culture of ethical data science practice
For those of us working for fair and accountable data systems, we have to think seriously about work and organizations if we want to think about the ethics of large scale data.
We believe that genuine collaboration between people who are working in data science and people who understand social theory can push for more ethical, and better, ways of knowing in increasingly data-saturated societies.
A key Mu3k theory made it into the @nytimes opinion pages this week. Touted and burnished by a professor at a top university WITHOUT transparency that Musk funded him for $10 Million for the research. How? 🧵
Humanity is collapsing from population decline. Sound familiar? It's the pet theory of ultra-rich Western white guys like Musk. He funded the center run by the author of this @nytimes op-ed with a $10 million grant. nytimes.com/interactive/20…
What I find EXTRAORDINARY that this funding is not listed on website for the center, the Population Wellbeing Institute, nor did the author say in his op-ed bio that he leads PWI. But Bloomberg broke this story last month about Musk's donation to PWI. bloomberg.com/news/articles/…
"If you do not have symptoms, you must not seek a test, as the scientific evidence shows that the test may not be able to detect whether you have the virus."
How in hell are we going into a second wave with this as the NHS messaging?
I say this having returned from a US state (KY) with free, rapid, on demand testing. Waste a test? Nobody wants to do nasal swab tests for fun. I flew back to the UK knowing I wasn't putting my family at risk (neg.)
And now I sit in 14 days legally mandated quarantine because I actually follow the rules.
The best minds in AI ethics are not even one step removed from this mess.
The goals here were clear: preserve standardisation in unprecedented times. But public trust? Fairness? Equity? People in charge ranked those goals lower than that of standardising marks across schools. But absence of 'ground truth' here -- 2020 individual exam results --> FAIL.
We need instead to understand how these systems are rolled out in practice -- not just open the so-called black box but what's around that box: who built it, who is using it, what are they doing with it, who do they think it is for, what do people know/THINK they know about it.
Data are for 2017 and cover the whole year. Of course we've only had 6 months of Covid data in 2020 (Feb-Aug), but already it it is the third top cause of death in the US. cdc.gov/nchs/data/nvsr…