Ai just fired you. now it wants to pay you $1,000 a month to feel less guilty.
They trained the model on your pull requests, your ledgers, your midnight translations. Two quarters later the same board that sang your praises hands you a cardboard box and a link to an NGO form. The severance? A 12-month, $1,000-a-month “AI Dividend” funded by the very code that replaced you. Fifty former developers, accountants and interpreters are already cashing the first checks; 3,000 more wait in a Google Sheet that smells of desperation and philanthropy.
The receipt for your obsolescence
AI Commons Project and What We Will scraped together $300 k from small donors and foundation crumbs. Their calculus is brutal: if a large language model can hallucinate a Python refactor faster than you can type it, the least the industry can do is tip the displaced worker like a barista. Kaitlin Cort, a software engineer turned activist, told me her recent boot-camp grads are now “human lint filters,” paid $18 an hour to proofread machine-born spaghetti. The gig lasts until the next release drops the error rate another two points. Then the lint filter burns, too.
The pilot is tiny, almost insulting. A grand a month won’t cover Oakland rent, let alone Madrid. But the NGOs are betting on narrative leverage: wave enough sad stories in front of OpenAI, Anthropic and Google and maybe shame mutates into a line item. Their ask is $3 million before New Year’s Eve—loose change for any firm that just spent $700 million on a training run. So far the silence is operatic.
Why the check feels like hush money
Universal-basic-income prophets—from Sam Altman to Elon Musk—love the concept until the invoice arrives. Altman’s own three-year UBI trial in Oakland proved people buy better groceries, not gym memberships. Meanwhile enrollment in U.S. computer-science programs has fallen for the first time since 2004; 19-year-olds watched GitHub Copilot autocomplete their homework and read the tea leaves. The major’s new reputation: arrive a dreamer, leave redundant.
Corporate PR departments recite the same lullaby: AI will augment, not replace. The layoff spreadsheets sing a different tune. Translation agencies now win contracts with per-word rates that only make sense if 80 percent of the job is machine output. Accounting boutiques brag about “AI-assisted closes” while quietly shrinking staff by a third. Each redundant human is a data point in a pitch deck: margin uplift 23 percent.
What happens when the dividend year ends? Cort admits most recipients will rotate through gig platforms, collecting micro-royalties from the same models that ate their salaries. A few enroll in retraining programs whose curriculums are still being written by—wait for it—generative AI. The loop devours its own tail and calls it upskilling.
Congressional offices have started requesting policy memos, but lawmakers are allergic to new entitlements during an election cycle. The EU’s AI Act mentions “worker adjustment funds” in Article 74, yet the text is so vague it could fund either a safety net or a hackathon. Lobbyists prefer the hackathon.
Meanwhile the price of compute keeps falling. Every $1 drop in GPU cost drags another profession across the automation event horizon. The NGOs’ $1,000 stipend is a moral photo booth: step inside, snap a picture of justice, post it, walk away. The line outside grows longer than the code waiting to replace them.
There is no regulatory cavalry on the ridge. Only the blinking cursor that finished your sentence before you did, and a payment portal that politely asks if you’d like the guilt money wired monthly or in a lump sum. Most choose monthly. It feels more like a salary, less like a payoff.