Existential Threat
“Existential Threat” is a key phrase in current A.I. discourse. The dialectic is between the idea that A.I. will be an existential risk to humanity in the near future, and the idea that it already is. The near-future group suggests we should all be scared of machines becoming sentient and trying to kill us, the already-here group contends that these machines are already actively harming people in much more banal, systemic ways.
A group of industry experts wrote an open letter to policymakers and news media (linked below), calling for better coverage of these issues:
“We reject the premise that only wealthy white men get to decide what constitutes an existential threat to society, and we call on policymakers and news media to diversify their sources. For people of color, women, LGBTQIA+ people, religious and caste minorities, indigenous people, migrants and other marginalized communities, technology has always posed an existential threat, it has repeatedly been harnessed to ensure our inferiority in societal power structures.”
At American Cyborg, we agree with the letter-writers here. Geoffrey Hinton saying his fears are more “existential” than Timnit Gebru’s is as absurd as it is cruel. It’s no surprise there is a gender divide; men have had forty years of techno-power-fantasy trips, women and non-binary people have had forty years of violent exclusion from technology’s development. The PBS report on Afghan asylum-seekers (linked below) is a timely example — machine translation is mischaracterizing the language in these applications, leading to our government automatically rejecting real people with real needs.
Large language models currently being deployed to deny people asylum, jobs, financial aid, and justice are a more pressing matter — and yes, more of an existential threat — than some distant, implausible Terminator scenario. The people who built this technology chose to ignore all the warnings along the way, and if we let them decide what we should all be afraid of while presenting themselves as our saviors, they’re going to built SkyNet just to prove themselves right.
Link List
Will A.I. Become the New McKinsey? by Ted Chiang, The New Yorker 2023/05/04
Ex-Googlers blast ‘AI godfather’ Geoffrey Hinton’s silence on fired AI experts by Wilfred Chan, Fast Company 2023/05/05
How language translation technology is jeopardizing Afghan asylum-seekers by Ali Rogen, PBS 2023/05/07
Global South AI Letter by Rosemary Ajayi et al, 2023/05/08
AI machines aren’t ‘hallucinating’. But their makers are by Naomi Klein, The Guardian 2023/05/08
We need to examine the beliefs of today’s tech luminaries by Anjana Ahuja, Financial Times 2023/05/10