To see the impact of the horrors these people endured is beyond words. These ugly companies are not trying to stop filth, they're probably teaching AI to make it. I pray their lawsuit is successful. God bless these people who stand up to evil!
It’s clear to see why billionaire oligarchs want to be rid of the middle class. They prefer we’re all working to death for peanuts. They think it’s the natural order.
It's truly absurd how few journalists even address the mechanical turk aspect of these AI companies. So many "AI" systems are just people identifying things and "training" the AI system that is being sold as if it already is doing the work.
it's so frustrating how overlooked this is in online arguments about ai. the priorities seem to be ip theft > illiteracy > environmental cost > human exploitation. it should be the other way around, and yet.
I think the biggest block for a lot of people in the western world is that we’re so divorced from how basics of nearly everything are produced. We receive and work with final products generally, but the production is outsourced so we don’t have to see it every day
If every American saw this it would immediately change the way we talk about AI. Thank you for sharing and for your integrity. Have a beautiful Thursday.
To a large degree, AI is a false front. The intelligence comes from Kenyans pounding keyboards to keep it all afloat. And they get no more credit than the people who pick produce.
Kenya is currently building a nuclear plant, the output of which might make it to Kenyans or it might go to whatever new digital coin is popular or into the AI farms.
The funny thing is that even in the US, this is happening. I was paid by Google for 5 months to curate their AI. It was for Gemini and Magi. All of us were through 3rd Party contractors and Google was Partnered with Global Logic. So I had 2 middlemen between me and the people I worked for.
Granted I was paid 22 dollars an hour. So it's not as exploitive. But I was also laid off one month after signing a extension to my contract. Tech companies are notorious for this and what used to seem like a dream job in the tech industry has turned into exploitation and lack of security.
The issue is external 3rd Party contractors don't get to join unions. They aren't considered "employees". The reason they set it up with multiple layers of middlemen and have you as a contractor is so you don't have the same benefits as the rest of the company's staff. Loopholes upon Loopholes.
Omg I use to work for remotask but it’s called Outlier now they definitely didn’t treat us like this. This is really sad cause that girl is right yall don’t do it here so what makes it ok to do in Kenya.
Isabel Wilkerson’s brilliant book describes the global caste system that supports - and requires - this kind of exploitation, and has warped social & economic relations for hundreds of years.
Time reported on similar exploitative labor practices with OpenAI in Jan 2023, but it didn't get much traction. Anecdotally when OpenAI and other AI services stop paying for labeling and filtering in ongoing effort, results become _terrible_. https://time.com/6247678/openai-chatgpt-kenya-workers/
I used to write for content mills, and many of my fellow writers were Kenyan. Now that AI has decimated that business, I'm guessing many are now training those same AIs...for a fraction of the (already terrible) wages.
A colleague and I wrote this piece on AI, energy, & Kenyan labor exploitation, but it didn’t start that way. We stumbled onto the labor part while researching & TIME magazine of all places was the one doing actual investigative reporting. Would love to see a place like ProPublica do a deep dive.
The channel "Digital Engine" on YT discusses cheerfully the progress of AI while at the same time bringing up the huge risks of AI becoming self-aware as it's being put in charge of our highest levels of decision-making and information. The sheer hubris of humanity is staggering.
Wouldn't that make the whole thing where you can tell if you're talking to an AI with certain trick questions not work? Like anybody who understands English well enough to type and respond to it and in it should be well aware there is no "i" in "arrangement"
If it exists, people have been exploited in its creation/refinement/everything. It’s a different form of human trafficking, & order for any of it to exist, humans have to ignore such crimes. And, yes: they are crimes.
I thought the mess that Outlier had for their US "hirelings" were bad...the 60 Minutes story is a complete horror show for the agents in Kenya. More needs to be addressed on the AI work farmed overseas, and the psychic and financial toll that's being placed on non-US workers.
Insane how much the wealth of the West is still to this day based on exploiting the rest of the world's resources and labour. And no one bats an eyelid.
The University of Virginia’s online conference “Rethinking the Inevitability of AI” yesterday had a panel on eugenics, race & colonialism in the AI boom. It’s key to disseminate this info to the public. Resource and digital extractivism have the same human cost. People don’t realize that AI = mining
Comments
Who could have guessed?
We definitely need to talk more about how AI attempts to replace "normal" jobs with exploitative PTSD-inducing gig labor in Global South.
https://internationalpolicy.org/publications/transnational-labor-policies-in-the-era-of-artificial-intelligence/
IDK why we expect people in countries with lower costs of living to endure the very worst of the Internet for an 8hr shift
https://www.curbed.com/article/citi-bike-battery-ebike-charging-docks-maintenance-repair.html
*and obvi Citi’s investment in climate destruction and genocide.
https://medium.com/palpable-voice/protecting-ai-from-humans-ae67ee2b9a1e?sk=a44522c4faf391b17864449cc98392f2
An evergreen quote by founder of readme, Gregory Koberger (2016):
How to start an AI startup
1. Hire a bunch of minimum wage humans to pretend to be AI pretending to be human
2. Wait for AI to be invented
https://medium.com/@tanyapobuda/a-strange-intruder-the-truth-about-artificial-intelligence-chatbots-ecf75003e6d4