Experts warn Carney’s AI strategy will lead to job cuts and environmental damage

Shortly before the House of Commons rose for the summer, Prime Minister Carney and AI Minister Evan Solomon released their long awaited and repeatedly delayed AI strategy, ‘AI for All.’ 

The strategy served as an update, and in some cases a duplication, of the preexisting Pan-Canadian Artificial Intelligence Strategy put forward by the Trudeau government in 2017; and comes just as venture capitalists, experts, and even some of the billionaire tech developers behind AI technology have begun predicting that the AI bubble may be about to burst.

The strategy, which includes $2B in federal spending on AI, comes at a time of unprecedented austerity in this country – with Carney divesting from human labour in favour of automation, ostensibly making our civil service more “innovative” and “efficient.”

This also comes on the heels of Elon Musk’s failed DOGE crusade in the United States which decimated their civil service, mined private data for yet unknown purposes, and led to hundreds of thousands of preventable deaths across the Global south.

“Setting a blanket goal of budget reductions for departments without an overview of the entire public service is anything but fiscally responsible. It is reckless,” PSAC head Sharon DeSousa said in reaction to Carney’s choice to follow Musk’s lead.

According to PSAC, the cost of these layoffs will be catastrophic: “People can expect longer wait times for passports, EI and childcare benefits, more unanswered calls at Canada Revenue Agency, reduced public health and food safety efforts, and a government that isn’t there for ordinary people when they need it most.”     

Like any workplace, there are roles in government that simply cannot be replaced by automation, but they’ve already started anyway. Public servants in departments and agencies across the country are deeply demoralized. 

One woman rabble spoke with in her late 60s who’d worked in government for two decades only to be fired last year without cause or explanation said she doesn’t know where or how she’ll ever find work again so close to the age of retirement.

Another public servant in her early 40s with two young kids whose husband works in government also tells me the “mood has been really low.”

“We’re losing huge talent here,” she said on condition of anonymity. “A group of my former colleagues got cut and I know about a few others. It just seems like this should be so clear and considerate for those affected and it’s not. The stress is building.”

Meanwhile, across the private sector, studies are already showing that 55 per cent of companies who fire large swaths of their staff in favour of AI eventually live to regret it.

In May this year, Meta let 8,000 employees go during an “AI-focused restructuring.” The move bred chaos and significant additional work for remaining staff, while destroying morale and degrading products. Worse, Meta reportedly made thousands of engineers train their own AI replacements before firing them. And now, mere weeks later, an internal memo obtained by Reuters says CEO Mark Zuckerberg admits he’s made “mistakes.”

Another tech giant forced to rehire  after an AI-inspired restructuring is Swedish e-commerce platform Klarna. Their CEO claimed he’d completely stopped hiring because “AI can already do all of the jobs that we as humans can do.” That same CEO would later admit that relying too heavily on AI led to poor customer service and falling satisfaction levels. Rising user frustration and an explosion in complaints eventually forced Klarna to rehire to restore service quality.

Lack of meaningful regulation

Director of the American digital rights advocacy group Fight for the Future, Evan Greer, says there’s nothing wrong with funding tech literacy education, but it is not a substitute for meaningful regulation.

“Focusing on ‘AI literacy’ assumes that a specific type of hyper scaling corporate AI is going to become dominant and everyone just needs to learn how to live with it,” Greer said in an interview over encrypted text.

“If we just accept that AI dominance is inevitable and all we can do is “educate” people, we’ve already given up the game. These technologies are built on a series of choices. Choices made mostly by very wealthy white men. We can make different choices.”     

Brain rot by robot 

In many workplaces where AI already exists, professional skills are literally degrading. A recent study from Wolters Kluwer, ‘Patients, doctors, and nurses on AI,’ shows a quantifiable ‘deskilling’ effect among accountants, software engineers, and practicing medical doctors who’ve made AI a routine part of their job. Some trainees have also been experiencing something called ‘never-skilling’ – whereby foundational learning opportunities to reason and think critically aren’t even happening anymore.

AI researcher and cofounder of Tech Reset Canada Jennifer Evans says tech regulations in this country are dangerously underdeveloped.

“It is actually obscene,” Evans said. “The federal government’s new “strategy” is … inadequate would be a kind word to use for it. I don’t think anybody really understands what they are governing or what the actual risks are.”

When Canada’s AI Minister Evan Solomon announced three new data centers in May, two of which planned for a densely populated part of downtown Vancouver, a journalist prefaced a question to Solomon by suggesting the executives at Telus – the multi-billion-dollar telecom giant leading the project – must be popping champagne bottles knowing they’re about to be rolling in cash.

Solomon responded with a variation of a line he often uses – that the Carney government is a “pragmatic” one, and therefore he is neither on “team pom pom” nor “team pitchfork” when it comes to artificial intelligence (the Telus CEO also claimed these data centres are going to be “sexy as hell,” which was interesting).

Days later, hundreds of would-be members of “team pitchfork” marched through downtown Vancouver carrying signs (not pitchforks) that said things like ‘F*** AI!’ and ‘WATER for plants, not techbro billionaires.’

In a phone interview from North Vancouver, 18-year-old protest organizer Torin Larocque said he was expecting up to 3,000 people at his next event. That second protest appears to have had an impact, and a petition opposing the projects has gained over 16,000 signatures as of this writing.

Several provinces eastward, another Canadian city and a similar protest. Residents of Hamilton, Ontario have been turning up at city hall by the hundreds to fight a data centre originally planned for their industrial waterfront.

Cadence Weapon, who lives just a few minutes away from the proposed site with his family, says he’s seen “the devastating toll” these can have on communities – so he’s been lending his fame to local protests. He’s also noticed AI plagiarizing his music and documented the experience on social media.

“There’s a sense that AI is here and data centres are inevitable. What’s happening now in Hamilton proves that isn’t true,” he said.

“And beyond these data centres, the tech they’re ushering in has become like a plague for real writers and artists. Yet Carney’s AI strategy made no mention of protecting the intellectual property of Canadian artists whose work is being scraped by AI algorithms without compensation.”

Not in my backyard

The growing public opposition to data centre development across the US and now Canada is not unfounded. People living near these centres have noticed worsening air quality, dangerous temperature spikes, and pollution to their local drinking water. Serious health problems have arisen as well from migraines to insomnia, to asthma and declining mental health from the constant hum of noise pollution.

Lyndsey Rolheiser, an assistant professor of economics at York University, recently co-authored the first comprehensive mapping of Canada’s data centre landscape. She discovered that commercial hyper scaling is indeed happening, but most of these projects are spatially concentrated and increasingly rural. They’re also all owned by foreign companies (American, mostly) which raises questions around data sovereignty and security. 

Perhaps somewhat predictably given their obsession with oil extraction, and despite a grid emissions intensity that is “nearly five times the national average,” Alberta accounts for over 90 per cent of planned data centre capacity in Canada. Just this week, Meta and the Government of Alberta announced another $13 billion data centre coming to the province.

“I think it is totally reasonable that the public is upset, because there is a lack of information,” Professor Rolheiser said during a phone interview.

“There’s such limited research on the benefits of data centres, at the local level. Like, how do they benefit a community? Other than potentially increasing property tax revenue.”

“We also don’t know the impact on local jobs. There’s limited direct employment at a data centre. But does it generate indirect jobs? That’s the big open question,” Rolheiser said.

An investigation conducted by PBS and Floodlight News into Elon Musk’s giant data centre development in rural Texas – where neighboring communities are largely made up of working poor Black families and retired military servicemembers who can’t afford to sue or relocate – has uncovered that both state and federal laws are being broken by this facility. There have been flagrant environmental violations, while residents are kept in the dark about expansion plans.

Here in Canada, the federal government has committed billions of spending to AI infrastructure (aka data centres), which the Canadian Centre for Policy Alternatives (CCPA) says is concerning without a plan to meaningfully regulate the industry. 

CCPA’s new report Not up to code: The potential downside of hyperscale data centres in Vancouver’s downtown says, “we seem to be relying on boosterism and wishful thinking that private corporations will prioritize the environment in their mad rush to cash in on AI infrastructure without asking the right questions about impact.” 

Jennifer Evans says the common thread here is that risk assessment is not being done by the Carney government; and if it is, it’s been shrouded in secrecy.

“There’s been no disclosure. We don’t know what, if any, risk assessment has been done, what that potential risk assessment has consisted of, or what it has told us.

We are regulating but in the worst way: regulating freedom and privacy as opposed to regulating the vendors of these technologies. And there’s a couple of high-risk vendors right now in addition to Palantir,” said Evans.

Follow the money

Much ink has been spilled on Palantir – the notorious data analytics firm known for it’s villainous exploits – but only recently have Canada’s connections to Palantir come to light. Federal contract agreements with the company were initially hidden and downplayed until a scoop by the Investigative Journalism Foundation revealed close to $50 million in spending on Palantir’s products and services.

Two big tech companies that the Government of Canada has long worked with (seemingly regardless of the financial cost or user experience) are Microsoft and Bell. Between 2021 and 2022, the feds spent an estimated $299.8M on contracts with Microsoft Canada, and an estimated $330 million on Bell Canada. Since Carney took office, there appears to have been an increase in funds flowing to both these companies.

Four months after his election, Carney signed an agreement with Canadian multinational tech giant Cohere, Inc. Now worth an estimated $7 billion, Cohere was purportedly tapped by the feds to “transform the public sector” with “secure, sovereign AI technology.” 

It’s unclear what this has entailed, exactly, but it appears as though the main thing Cohere does is roll out chatbots – presumably across the federal departments where mass layoffs are happening. In a recent interview, author of the book Enshittification, Cory Doctorow, argued that not only is it impossible to replace our civil servants this way, doing this now is both “crazy” and “objectively very stupid.”

A statement on Cohere’s website reads as follows: “Accelerating AI adoption will deliver massive productivity and efficiency gains to enhance public services and modernize operations. This partnership will help drive Canada’s leadership in AI forward, strengthening economic growth and national security.”     

In June, Bell Canada and Cohere announced they’d be partnering on a major AI infrastructure deal. From the limited information available publicly, it seems Bell will be assisting Cohere by letting them use their preexisting data centre supply.

Cohere is led by Aiden Gomez, who says he wants to “save the world with AI” and make it “a part of everybody’s workday.” He’s also predicted that AI will allow us to automate “any human task” within five years. AI Minister Evan Solomon, who worked as a journalist before entering politics (never directly in tech), believes Gomez and his colleagues are “geniuses.”

One comment from Aiden Gomez that stands out: “Don’t believe a lot of these stories that have also fueled stories of terminators and doomsdays and these sorts of sci-fi narratives that have emerged. They’ve since become unpopular, because people have been faced with the reality of the technology. They get to watch and experience it themselves.”     

Contrary to the hype, AI labs like Cohere are known to be massive fiscal loss-makers. The revenue their technologies generate, relative to the obscene capital needed to develop and sustain them, seldom delivers return on investment.      

The ‘last frontier of colonization?’

In her bestselling book, Empire of AI, tech journalist Karen Hao shares two radically divergent visions of what AI can be and for whom it can work. 

She reflects on an investigation conducted in New Zealand, where she discovered a Māori researcher and his partner who’d developed an artificial intelligence for the sole hopeful purpose of reviving his Indigenous language. Their work was approached slowly and methodically, with the complete informed consent and permission of Elders and Indigenous community from which they’d be training their model. 

“Data is the last frontier of colonization,” Hao was told by the developer. Data stemming from sacred oral histories and Indigenous languages like theirs are the last resources the billionaire class can extract from humanity. Therefore, this data must be protected at all costs. Hao juxtaposes this approach to the violent extraction normalized in Silicon Valley through companies like Sam Altman’s Open AI – where to forcefully impose, to ‘move fast and break things,’ is expected and even celebrated.

Hao wrote, “What I reject is the dangerous notion that broad benefit from AI can only be derived from – indeed, will ever emerge from – a vision for the technology that requires the complete capitulation of our privacy, our agency, and our worth, including the value of our labor and art toward an ultimately imperial centralization project.”

In the same city where AI Minister Evan Solomon announced two new data centres in May, neighbouring residents are currently under mandatory water use restrictions – the strictest in over a decade. Citizens of Metro Vancouver are forbidden from watering their own lawns or filling their swimming pools.

A single data centre uses up to 19 million litres of water a day.

The post Experts warn Carney’s AI strategy will lead to job cuts and environmental damage appeared first on rabble.ca.

Related Products