TeaAI ethicist Eleanor Dredge believes that in order to move forward with artificial intelligence, humans need to recognize the “humanity” of AI. He means that while we often talk about AI as if it were some mysterious, formless thing, it is the product of hours of human labor in silicon and quartz mines, microchip factories, and data labeling centers. If citizens want to wrest power from the tech companies that build AI, it helps to clearly understand what they are selling us. It also means seeing it through marketing and promotion: “The cloud,” that obscure, distant place where your photos are stored, is just someone else’s computer; The AI ”hallucination” is a system error or data labeling mistake; Intelligence is a very broad – and potentially misleading – term.
The story in Silicon Valley is that AI will either save us, ushering in a new era of prosperity and human longevity, or it will wipe us all out. Drèze argues, convincingly, that big tech cannot possibly fulfill its utopian promises as long as it continues to pursue profit and power above all else, and while it serves the interests of a narrow few. Meanwhile, the cultural obsession with AI-induced apocalypse easily diverts attention away from practical conversations about making AI safer — by ensuring that citizens have more power over how their information is used, and how AI models are trained and regulated, for example. The apocalyptic vision of tech titans also creates the false impression that AI security is more important than climate change, geopolitical tensions and inequality, three problems that are already causing large amounts of suffering and that, if addressed, would also reduce many of the risks associated with AI.
Dredge, an assistant research professor at the University of Cambridge’s Leverhulme Center for the Future of Intelligence, has conducted research that refutes AI-driven recruitment processes, and highlights the dangers of incorporating AI into law enforcement. Many of his criticisms of big tech will sound familiar: that AI algorithms perpetuate and reinforce pre-existing biases; Despite all their vapid, generic promises to be “transparent” and “accountable,” tech giants have become too big to care about ethics; A handful of officials have acquired too much power; And that the rollout of AI is happening without our consent and without consideration of planetary health. But his writings have intellectual value, and his philosophical explorations of, for example, how we define intelligence and why it matters, are very interesting.
At times, however, Dredge’s arguments become hasty and convoluted. “It is merely an illusion that the labor force is disappearing: work is simply shifting from artists to data labelers and from marketers to engineers,” she writes at one point. But it is not clear that artists and “marketers” will be made redundant in the coming AI age. It’s not as if data labelers working for poverty wages in Kenya would have been artists 10 years ago. Equally, given the due importance she places on promoting diversity in AI, it is disappointing that she dismisses unconscious bias as “just another way of taking the perspective of the perpetrator rather than the recipient of harm”, as if intent is completely irrelevant. She writes that while people are ready for “less quiet” conversations about equality, I would add that these conversations also demand precision and nuance.
The title of the book suggests that Dredge wants to present an alternative vision of what AI could be. She gives some examples of the kind of ethical, community-based AI projects she believes should replace top-down big tech products, and many of them are strangely weak. These include an app called Mumkin, designed to facilitate difficult conversations about female genital mutilation (FGM) in India, and Kuini, an AI chatbot designed to help Māori women quit smoking. In both instances I found myself wondering what value AI brings to these efforts – especially given the discussion Dredge has had about its environmental impact and exploitative supply chains. Is an app more effective than training community members to act as mentors, or funding support groups and peer support workers?
Dredge describes the toolkit he co-designed to help AI companies act ethically and comply with EU regulations as a “cornerstone of utopianism”, as it focuses on how AI is created and created. In his view, utopia is not a destination but “a process of working together to develop a fairer world”. Perhaps it would be unreasonable to expect a very specific utopian vision, but I would still be interested to know how AI, in particular, could help us create a better future. She doesn’t mention things that many of AI’s biggest critics nevertheless welcome, such as its ability to improve drug discovery and disease detection. This book asks what would happen if we got AI right? After reading this, I can’t say I really know.
After newsletter promotion
