The astonishing amount of technological and financial resources that, this decade alone, have been poured into the development of autonomous automobiles indicates the market's total commitment to maintaining the profitability of a form of transportation that is in every way inefficient. So, we have on one side, the Green New Deal, which wants to make huge public investments in high-speed rail; and on the other, a market solution to the dangers of driving, the non-stop jams, the long list of negative externalities (many of which are contributing to climate change) that amounts to nothing more than replacing one luxury with another.
The first luxury is the ownership of a car, which, for much its time, does nothing but wait for the driver. The second is a car that drives the owner. The luxury associated with the latter goes all the way back to the aristocratic horse-drawn carriage. The robot becomes your coachman. You become a gentleman or lady. The transition from one luxury to another (carriage to carriage with coachman) will, however, not solve any part of the current transportation crisis and the deteriorating vehicle infrastructure. The jams will remain, and also the dangers. But there is something more: A recent study discovered that autonomous cars present a greater danger to pedestrians with dark skin than those with light skin.
If you’re a person with dark skin, you may be more likely than your white friends to get hit by a self-driving car, according to a new study out of the Georgia Institute of Technology. That’s because automated vehicles may be better at detecting pedestrians with lighter skin tones.
The Vox post points out two important things about the study, “Predictive Inequity in Object Detection,” conducted by Georgia Institute of Technology. One, it only had access to datasets from academic research because corporate firms investing in our robot-car future refuse to share key or useful information with the public. The other concerns the racial flaw of the robot cars. Is this really about the machines? Is it just hard for them to detect dark skin? Should black people wear bright clothes or masks to help the robots recognize them as walking humans?
It seems, according to Vox, that robots are actually just fine. What prevents them from recognizing a black pedestrian as well as white one are the brains for their brains—their programmers, many of whom are white and work and live in very white environments, and as a consequence do not automatically visualize (or model) a person with dark skin when coding technologies that closely interact with human bodies. This results in the "problem of algorithmic bias."
Since algorithmic systems “learn” from the examples they’re fed, if they don’t get enough examples of, say, black women during the learning stage, they’ll have a harder time recognizing them when deployed.
And now, two things to think about.
One: "Pedestrian fatalities are soaring in the U.S. and are expected to reach their highest level in 28 years."
Two: "Report finds $23 billion racial funding gap for schools."
Just last week the @washingtonpost reported on a *$23 BILLION* racial funding gap in K-12 education. White school districts received $23b more than nonwhite districts.
- Mass Incarceration
- Banks targeting Black Americans before the housing crisis (via @guardian) https://t.co/WmW9wwKlg0
— Alexandria Ocasio-Cortez (@AOC) March 7, 2019