9 hours ago
9
Bad news, baby. The New Yorker reports the rapid advance of AI in the workplace will create a “permanent underclass” of everyone not already hitched to the AI train.
The prediction comes from OpenAI employee Leopold Aschenbrenner, who claims AI will “reach or exceed human capacity” by 2027. Once it develops capacity to innovate, AI superintelligence will supersede even a need for its own programmers … and then wipe out the jobs done by everyone else.
Nate Soares, winner of “most sunshine in book title” and co-author of AI critique If Anyone Builds It, Everyone Dies suggests “people should not be banking on work in the long term”. Math tutors, cinematographers, brand strategists and journalists are quoted by the New Yorker, freaking out.
The consolation here is that if you are among those panicking about being forced into the permanent underclass, you are already in it. Inherited wealth makes more billionaires than entrepreneurship, the opportunity gap is growing; if your family don’t have the readies to fund your tech startup, media empire or eventual presidential ambitions, it’s probably because they were in a tech-displaced underclass, too.
My own family were agrarian stablehands before industrialisation drove them to work in the Glasgow shipyards. After industrial decline there, they migrated south into retail clerkery. My dad worked in betting shops that have been replaced by apps. My mum was an admin assistant whose tasks you can now run from an iPhone.
So if this column is replaced by the burps of a machine programmed to deliver cultural trend-based leftist snark, I’ll be following family tradition. Spinning jennies, combine harvesters, electrification, trains, cars, recording, computers and, presumably, the wheel – given visible lack of sedan chairs – have all devastated industries and with them, people.
How well we cope with that devastation is the opportunity of the hour, and you don’t need a machine to conclude unionisation is the remedy.
Pals, we’ve been here before.
When automation hit the Japanese carmaking industry in the 1970s, its unions successfully fought for “cooperative modernisation”, with job security guarantees and lifelong retraining programs.
With technological progress replacing many of their functions, throughout the 1980s Australian nurses fought attacks to reduce their status and instead bargained to enhance their productivity through tertiary training with the new technologies – a victory that improved both their work conditions and professional standing.
We are already seeing industrial unions take the fight to AI. The Writers Guild of America strike in 2023 resulted in an agreement that determined writers wouldn’t be replaced by generative AI as source material, or be compelled to use it – but would be at liberty to use AI as a productivity tool at their own writerly discretion, compliant with company policies.
The opportunity to stand up for our AI-threatened livelihoods is now, because – despite the relentless insistence of tech’s grifters, zealots and cheerleaders – AI is not industrially inevitable. At present, it’s not even sustainable.
While AI data centre construction skyrockets, there are costs emerging even beyond the energy and water required to power and maintain them. Harris “Kuppy” Kupperman from hedge fund Praetorian Capital calculates the physical depreciation of these centres may be as rapid as three years. “No one understands,” was the quote collected by Futurism, “how the financial math is supposed to work.”
Not investors, with the majority of responders to a Bank of America survey admitting concern that AI stocks were in a bubble. Not the Bank of England, who’ve warned of a potential “sudden correction” in worldwide markets as it raised concerns about the soaring company valuations.
Not MIT, whose own maths have revealed that 95% of corporate clients trialling AI pilots or initiatives find zero – or negligent – return, or Carnegie Mellon University, who released a paper showing AI agents failed to complete basic office tasks about 70% of the time. Not TechRepublic, reporting that 55% of surveyed companies that sacked workers to replace them with automation now regret the decision. Analytics firm Orgvue also found that layoffs cost the companies $1.27 for every $1 saved by the sacking.
It sure makes the claim by professional tech patsy Nick Clegg that paying copyright on AI’s stolen content through a model such as a local library can manage would “basically kill” the industry revealing.
Not paying for a product you exploit to sell your product is theft; not paying for labour you employ is slavery. If your business model is theft and slavery, you’re not an industry. You’re a pirate. Perhaps this is techno-cynicism. Or perhaps, this week, OpenAI CEO, Sam Altman, announced a ChatGPT upgrade with tits. Ahoy! It’s some comfort that those who would casually erase our jobs at least cater to our entertainment.
For my amusement, I used AI to create and automate a “techbro bot” that generates social media posts as a techlord, informed by a dataset of public statements by same.
My AI couldn’t quite manage the code, so I had to recruit a professional, but I stare at its vapid, techo-capitalist self-hype on the hour, and I marvel how indistinguishable, digitisable – and replaceable – the makers of this bubble really are.
-
Van Badham is a Guardian Australia columnist
