While AI Companies Promise Heaven, Underage Workers Are Training Their Models

The rampant cynicism in Silicon Valley AI companies is beyond me

Perhaps the most startling reality of AI, in the worst possible sense of the word, is that in the same week, you can find an article entitled OpenAI chief seeks new Microsoft funds to build ‘superintelligence’ published next to another entitled Underage Workers Are Training AI.

A profoundly dystopian image emerges out of the sheer contrast between these two perspectives of the same AI revolution.

In one article, you get Sam, repeating his notoriously visionary claims about AI being a catalyst for an abundance of intelligence. OpenAI’s product is not ChatGPT (that’s just a trifle) but intelligence — and freedom of choice — for all.

On the other, you get Hassan, an eighteen-year-old Pakistani who started working in data labeling at fifteen as a gig worker doing outsourced tasks for more than 10 hours/day to reinforce the right behavior in AI systems — which surely will, in no time, free him of his predestination, as Sam predicts.

While Sam, the leading CEO of the AI revolution, promises the “magic intelligence in the sky,” Hassan, the forsaken worker of the invisible AI factory, says he lives under “digital slavery.”

That’s the greatest marketing operation that AI companies have successfully executed: Making people believe that AI is the engine that will take us all, together, to the stars when in reality it is deeply rooted — and rotten — into the worst vestiges of a world that we should have left behind long ago.

Like Saiph Savage says, “People just simply don’t know that there are human workers behind the scenes.” No matter how many times it is reported that this practice is commonplace and silently embraced by all major players, it always ends up buried — crushed in between articles praising ChatGPT’s feats, manifestos about the undisputable benefits of innovation, and open letters on the unfathomable risks that await us in an unpredictable future.

Now, of course, the platforms “require that workers be over 18” but these underage kids easily find ways to bypass the checks, driven by pure basic survival necessity. That’s the direct fault of the crowdsourcing platforms, but the AI leaders we all know so well, including OpenAI, Google, Anthropic, and Microsoft, are not free of responsibility. Here’s what Niamh Rowe writes for Wired:

Crowdsourcing platforms such as Toloka, Appen, Clickworker, Teemwork.AI, and OneForma connect millions of remote gig workers in the global south to tech companies located in Silicon Valley. Platforms post micro-tasks from their tech clients, which have included Amazon, Microsoft Azure, Salesforce, Google, Nvidia, Boeing, and Adobe. …

These workers are predominantly based in East Africa, Venezuela, Pakistan, India, and the Philippines—though there are even workers in refugee camps, who label, evaluate, and generate data. Workers are paid per task, with remuneration ranging from a cent to a few dollars—although the upper end is considered something of a rare gem, workers say.

The cynicism that surrounds everything going on in AI is terrible. For years, experts have claimed that as AI gets better, it grows independent of human input. For years, generative AI advocates have defended that it’s a vehicle for democratizing art and creativity. For years, executives of AI companies, especially Altman himself, have promised that AI will create a post-scarcity society where the only problem we’ll have to face is how to distribute the immense amounts of wealth that will result from the technology.

For years, that discourse has dominated the AI spaces and it’s now conditioning policy and public opinion. For years. While at the same time, day after day, day and night, in some forgotten parts of the world — a world we all share but most days it doesn’t seem like it — thousands of miserable youngsters had to go, painstakingly and painfully, through a deeply psychologically-damaging process just to earn a livelihood.

And I ask myself, are the companies that ignore this reality while being complicit in it, the ones we want to be the creators of that promised land?

How can they claim, unabashedly, that AI will be the greatest invention ever, more impactful than fire and electricity, and more beneficial than medicine and education while knowing this is happening under their feet and behind our backs? While knowing that their beloved AI is being erected on the sweat and tears of the most vulnerable people. There are no words, really.

I want to end this rant — which shouldn’t be a rant but a banner beside the logo of each and every AI company that engages in this shameful practice — by repeating something I published earlier on Notes:

This invisible reality deeply and hopelessly stains all the efforts in AI that I’d otherwise applaud. I don’t think this criticism applies to everyone in the space (I truly don’t think so), but it’s an ethical burden that will sink the whole ship in the eyes of the world.

(This article has no image. I wanted to make one but felt out of place. While I was trying, I asked DALL-E to transform this essay into an image. He told me that “The themes present in the document are complex and sensitive, involving real-world issues that require careful handling.” It may seem that making this issue overt is more damaging than carrying it out in the first place. Shameless. Also, I now know that the only reason DALL-E can tell me that the issues are “complex and sensitive” is that someone, somewhere, probably far away from here and possibly underage, had to teach him once, twice, three times, for hours on end, what is sensitive and what is not - what underworld reality he is allowed to reflect and what must remain hidden).