The idea of a worse future due to AI is rooted in understanding how automation of bad things happened in the past (e.g. philosophy, politics, economics), and how that will mean even more bad things ahead.
Ifeoma Ajunwa has written an amazing article giving exactly the kind of example we all should study.
He points out what any intelligent human operator would key in on: known qualifications.
…as Sen. Amy Klobuchar helpfully shared on the first day of hearings for Judge Jackson’s confirmation, the nominee has more judicial experience than “four people who are already on the Supreme Court.”
And before that, he points out what un-intelligent (e.g. dumb robot) operators get stuck on: fear based on the unfamiliar or unknown (e.g. racism).
If there were any doubt that this social discrimination still exists, just consider the minimally veiled racist remarks Tucker Carlson made about Judge Jackson’s African name while questioning her credentials. “So, is Ketanji Brown Jackson—a name that even Joe Biden has trouble pronouncing—one of the top legal minds in the entire country?” he asked. In the world of Fox News, an African name alone is disqualifying.
So true. So well said.
But it gets even better. Ifeoma links his analysis to a history of systemic racism in America.
In 2004, Marianne Bertrand and Sendhil Mullainathan published a now-famous field experiment they conducted to test racial discrimination in hiring. They responded with fictitious résumés to help-wanted ads in Boston and Chicago newspapers. They assigned each résumé either a very African American–sounding name (think Lakisha and Jamal) or a very White-sounding name (think Emily or Greg). One result: In general, résumés with White (read European) names received 50 percent more callbacks for interviews. Bertrand and Mullainathan also found that the amount of discrimination was the same across occupations and industries.
The employers did not know for certain that the job applicants were Black Americans—they were making the conclusion based on how they perceived the name.
How our brains perceive a name can be actively manipulated by “deliberately misleading” partisan extremists like Tucker “gas chamber” Carlson.
For weeks, Fox News host Tucker Carlson’s prime-time show has featured racist attacks…
The fact that an alleged asset of foreign military intelligence (Russia) can poison information so openly in America is disappointing on its own, of course. Even worse is when automation technology gives them far more impact.
Fox “news” is basically this:
…open-ended endorsement of white supremacist ideology…
Recognize that Fox is a past form of automation and therefore foreshadowing of the ugly side to AI. The ability to move racism faster and wider, to create more harm more quickly than before, will be the outcome of “open-ended” use of technology for hate.