Remember the movie Soylent Green? It’s a dystopian classic set in 2022. The government provides the masses rations of Soylent Green to eat. Today’s artificial Intelligence has the same dirty secret that the film’s hero, Robert Thorn, a detective with the NYPD, uncovers in a grand scene.
Overpopulation has devastated the earth’s ability to sustain us. Millions of souls are jammed into megacities like New York. Hunger and thirst are daily companions. The government desperately tries to respond with supplemental food rations but is struggling. Euthanasia is encouraged and use of “going home” services and centers makes is easy.
Thorn’s good friend and roommate, Sol, had been helping him in a murder investigation of a powerful industrialist who is also a director of the Soylent Corporation. The firm’s product is Soylent Green, the free food ration supposedly made from ocean algae. Sol is driven to despair when he discovers in a secret report that the oceans are actually dead and for years the ration has been made from some thing else – but what?
Sol, elects to “go home” and dies in a lovely chamber while wall size images of the earth, when it was full of life, plays upon IMAX wall size screens and a medley of beautiful masterpieces by Tchaikovsky, Beethoven and Grieg surround him in sound. During Sol’s final moments he tells Thorn the truth about Soylent Green and begs him to follow the handling of his corpse. Inside a large processing plant Thorn discovers what is really going on. He screams out in anguish: “Soylent Green is people!” And, that is also A.I.’s dirty little secret: A.I. is made from people!
We are not talking about the six and even seven figure compensated guru’s who devise the deep learning and machine learning algorithms. No, A.I. is critically dependent upon a vast base of hundreds and thousands of low paid grunts. The AP broke the story last year. Let’s hear from the author Ryan Nakashima’s introduction:
“From makeup artists in Venezuela to women in conservative parts of India, people around the world are doing the digital equivalent of needlework —drawing boxes around cars in street photos, tagging images, and transcribing snatches of speech that computers can’t quite make out.
Such data feeds directly into “machine learning” algorithms that help self-driving cars wind through traffic and let Alexa figure out that you want the lights on. Many such technologies wouldn’t work without massive quantities of this human-labeled data.
These repetitive tasks pay pennies apiece.”
Beyond the ethically dubious pay rates consider the other consequences. Humans – even in support of artificial intelligence – bring their biases, feelings and errors with them. Why do you think Google’s hate speech detecting A.I. is biased against black people? Accurate people driven labeling could make the difference between a self-driving car distinguishing between the sky and the side of a truck — a distinction Tesla’s Model S failed in the first known fatality involving self-driving systems in 2016.
And as we know A.I.’s are just not up to the skill level of humans yet. When they reach their limit what do they do? Turn to humans for help. Facebook’s 15,000 content moderators spend hours reviewing vile pornography, hate speech and terrible violence because A.I.’s cannot determine if the content is allowable. Humans aren’t built for this and many are now suffering PTSD (Post Traumatic Stress Disorder) due to the day-in, day-out trauma they have experienced.
Investors have recognized the market opportunity and funds are flowing into this sector. In addition to the traditional Indian IT outsourcing firms who pioneered exploiting offshore labor (like, Cognizant who does Facebook’s content moderation) there are venture-backed start-ups like Mighty Ai who claim over 300,000 contractors world-wide. Venture capitalist S. “Soma” Somasegar says he sees “billions of dollars of opportunity” in servicing the needs of machine learning algorithms. His firm, Madrona Venture Group, invested in Mighty Ai. Humans will be in the loop “for a long, long, long time to come,” he says.
The whole thing seems kind of sick, doesn’t it? After much hype and years of promise artificial intelligence suddenly broke out of its “AI Winter” and seems to be everywhere. But, because of its limits it rests upon a foundation of hundreds of thousands of poorly paid people, some obligated to review the sickest outputs of our humanity. We are serving the needs of the machines. How’s that for an example of the “Law of Unintended Consequences”?
By John Pientka