How big data is unfair

Moritz Hardt, Medium, Jul 31, 2015
Commentary by Stephen Downes

Is justice 'fairness'? Is there a requirement that big data be fair? That is the underlying presumption behind this paper that argues that the needs and interests of minorities are subsumed under the unflinching generalizations of big data. Empirically, I think there's no doubt that Moritz Hardt is right. This is the sort of observation that has spurred philosophers since John Stuart Mill to warn of the "tyranny of the majority". How much does it matter, though? Will it even slow down the adoption of big data? It should - but will it? In medicine, we have the "do no harm" principle to prevent doctors from unthinkingly prescribing stock solutions to special cases. But we have no equivalent in education. We don't really get an answer - and at the very end I see the purpose of the article is not to actually address the issue, but to promote a conference. How disappointing.

Views: 0 today, 389 total (since January 1, 2017).[Direct Link]