Hardcover, 312 pages
Published by MIT Press.
As the authors say data does not flow to the lowest point like water it spreads out reaching unexpected places. They look at how even data collected for the best reasons can, and often does, get used in unintended ways. I appreciated that they don’t try to present solutions but do speak to how things can get better, slowly. Highly recommend if you are at all interested in data collection, privacy, or related topics.
I was lucky to get an early look at this (it's out in April) for a review I'm writing. A number of interesting key ideas in the book, which aims to offer a broad account of "the secret life of data" without falling prey to the many "bullshit" accounts from the tech press industry. The authors succeed in this regard.
The secret life of data is premised on this idea:
“There is no limit to the amount and variety of data - and ultimately, knowledge - that may be produced from an object, event, or interaction, given enough time, distance, and computational power” (xii).
They also develop the idea of "algo-vision":
“The widespread and disorienting experience of seeing oneself through the ‘eyes’ of the algorithm” (xx)
Finally, they offer the notion of "triangulation" as an ethical approach to tech development
“A model for artificial …
I was lucky to get an early look at this (it's out in April) for a review I'm writing. A number of interesting key ideas in the book, which aims to offer a broad account of "the secret life of data" without falling prey to the many "bullshit" accounts from the tech press industry. The authors succeed in this regard.
The secret life of data is premised on this idea:
“There is no limit to the amount and variety of data - and ultimately, knowledge - that may be produced from an object, event, or interaction, given enough time, distance, and computational power” (xii).
They also develop the idea of "algo-vision":
“The widespread and disorienting experience of seeing oneself through the ‘eyes’ of the algorithm” (xx)
Finally, they offer the notion of "triangulation" as an ethical approach to tech development
“A model for artificial intelligence and machine learning systems…based on a multiperspectival approach to knowledge creation as a more equitable and accurate alternative to the single-point perspectives presented by many platforms.” (xx)
My favorite part of the book is its argument that collective approaches will have to beat out individualized ones if we are to actually deal with how much data is out there and how many nefarious actors are champing at the bit to use it, aggregate it, make money from it, etc.
