The Guardian said Eleanor Drage’s What If We Got AI Right? argues that AI is built on human labour in silicon and quartz mines, microchip factories and data labelling centres. That’s the machinery behind the glossy sales pitch. Not magic. Not inevitability. Work, extraction and a lot of corporate theatre.
The Labour Behind the Machine
The review said Drage wants humans to recognise AI’s "humanity," by which she means the hours of labour buried inside the system. Silicon and quartz mines. Microchip factories. Data labelling centres. The cloud, the review noted, is just someone else’s computer. The miracle dissolves fast when you look at who does the work and who owns the thing.
Drage, an assistant research professor at the University of Cambridge’s Leverhulme Centre for the Future of Intelligence, has researched AI-powered recruitment and the dangers of AI in law enforcement. The review said her criticism runs through the usual corporate fog: AI algorithms perpetuate and entrench pre-existing biases, tech giants have become too big to care about ethics, and a handful of executives have amassed too much power. The language is polite. The picture isn’t.
The review also said Drage argues that AI is being rolled out without consent and without consideration for planetary health. That matters because the people making the decisions are not the people living with the consequences. The companies sell speed, scale and inevitability. Everyone else gets the bill.
Brussels Rules, Corporate Power
Drage described a toolkit she co-designed to help AI companies act ethically and comply with EU regulations as a "cornerstone of utopianism," because it shifts the focus to how AI is built and made. That’s where the real fight sits, according to the review: not in the marketing copy, but in the architecture of the system itself. The EU regulations appear here not as a democratic cure-all, but as another layer in the management of corporate power.
The review said Drage argues that if citizens want to wrest power from tech companies building AI, they need to understand what those companies are selling and see through the marketing and hype. That’s a blunt admission of how the game works. The product is not just software. It’s influence, dependency and control dressed up as progress.
The review said Drage believes the cultural obsession with AI-induced apocalypse distracts from practical conversations about making AI safer by giving citizens more power over how their information is used and how AI models are trained and regulated. That’s the real terrain: who gets to decide, who gets to train, who gets to profit, and who gets to refuse.
Who Gets to Call It Utopia
Drage wrote that "It’s merely an illusion that the labour force is disappearing: work is simply shifting, from artists to data labellers and from marketeers to engineers," though the reviewer questioned that claim. The line still lands where it hurts. The labour doesn’t vanish. It moves. The people at the bottom keep doing the work, only now it’s hidden behind cleaner branding and a more expensive vocabulary.
The review said the book includes examples such as an app called Mumkin, designed to facilitate difficult conversations about female genital mutilation in India, and Kuini, an AI chatbot built to help Māori women quit smoking. Those examples sit inside the book’s broader argument that utopia is not a destination but the "process of working together to develop a fairer world." Nice words. The question, as ever, is who gets to define fair, and who gets paid while doing it.
The reviewer said Drage does not mention AI’s potential to improve drug discovery and disease detection. That omission matters in a review that otherwise tries to separate hype from reality. The final line left the whole thing hanging: "What if we got AI right," the review asked, and after reading the book, the reviewer said they could not say they really knew.
That uncertainty is the most honest thing in the piece. The industry promises salvation. The book, at least as reviewed here, points back to labour, regulation, bias and power. The rest is just the usual corporate fog, with a few more servers humming in the background.