Ross: AI is reliant on exploitative outsourced labor too
Sep 5, 2023, 8:05 AM | Updated: 8:42 am
FILE - The OpenAI logo is seen on a mobile phone in front of a computer screen displaying output from ChatGPT, March 21, 2023, in Boston. The U.S. Federal Trade Commission has launched an investigation into ChatGPT creator OpenAI and whether the artificial intelligence company violated consumer protection laws by scraping public data and publishing false information through its chatbot, according to reports in the Washington Post and the New York Times. (AP Photo/Michael Dwyer, File)
Credit: ASSOCIATED PRESS
(AP Photo/Michael Dwyer, File)
A story in the reveals the little-known secret behind artificial intelligence. It relies on a workforce of hundreds of thousands of humans to tell the machines what they鈥檙e seeing.
I鈥檓 embarrassed to say I had assumed that today鈥檚 AI software was smart enough to categorize objects all on its own.
More from Dave Ross: A perfect match of willful blindness and Trump’s trial
But according to the Post, these software brains are totally dependent on the actual brains of a beehive of young overseas gig workers tapping away at crowded offices and Internet cafes all over the world.
And what are they doing?
Annotation.
They annotate still photographs and driving videos, labeling objects so the computer knows what to call them. They鈥檙e the reason a computer can differentiate a pedestrian from a snowman or a fire hydrant from a small child (should the self-driving car have to make a tough choice in an emergency.)
The focus of the article is how these workers are exploited in the usual ways: low wages, no standard work rules, no government oversight, and that鈥檚 wrong, of course. Exploitation is bad.
But what鈥檚 just as alarming is that all this amazing computer intelligence is totally dependent on these exploited workers who are racing through images as fast as they can and don鈥檛 always feel valued by their employers.
The article mentions one annotator who spent three days on a job that was supposed to pay $50 but only got $12. Another worker spent four hours on a job hoping to earn $2 but ended up with $0.30.
I鈥檓 sure these workers understand the consequences of making a mistake, but what if they鈥檙e mainly there for the money and aren鈥檛 being particularly careful? Or they鈥檙e angry about being ripped off?
You might wonder, as you are being whisked along by your driverless car, if someone tapping away in a coffee shop on the other side of the world just might have been disgruntled enough to mistake a pedestrian for a snowman.
Listen to Seattle鈥檚 Morning News with Dave Ross and Colleen O鈥橞rien weekday mornings from 5 鈥� 9 a.m. on 成人X站 Newsradio, 97.3 FM. Subscribe to the聽podcast here.
