An AI apocalypse isn’t the problem – technology-driven inequality is
Fears of an artificial intelligence apocalypse make the news, but it’s AI-fuelled inequality we should worry about, says Andrew Simms
Josie Ford
By Andrew Simms
ONE of the biggest potential impacts of artificial intelligence is often overlooked. Rather than the frequently touted extremes of technological utopia or an end to humanity, AI could entrench and deepen the status quo, intensifying business as usual by ramping up overconsumption and inequality. For many scientists, this is a big concern.
Scientists for Global Responsibility, a campaign group for scientists and engineers that I work for, recently surveyed its 750 members about AI’s effects on the future. Nine out of 10 respondents thought that AI would deliver more power and economic benefit to corporations than to citizens. Eight out of 10 thought AI would lead to a dystopian future, rather than a utopian or unchanged one.
Mark Carney, governor of the Bank of England, recently said AI is part of a fourth industrial revolution, which will not only tilt the balance of power further away from low-paid workers to the owners of capital, but “substantially boost productivity and supply”. In other words, AI will enable us to make a lot more stuff using fewer people, and as a result is likely to worsen overconsumption and unemployment levels.
Predictions about how AI and automation will affect work suggest that anything from 35 to 50 per cent of all jobs could be at risk, according to the University of Oxford and the Bank of England. Profits from this change are likely to flow to corporations and their owners rather than their workforces, deepening inequality and exacerbating the decline in wages relative to global wealth seen over the past 20 years.
Alarm bells about an AI dystopia are already ringing. There are fears about the development of weapons that decide for themselves who to kill – so-called killer robots – and about how AI could lead to a supercharged surveillance society, where everything you do is tracked and recorded. AI is also being used to intensify environmentally damaging resource extraction. An embattled oil and gas industry sees AI as a “godsend”, as one leading industry journal put it, and is now using the technology to help find new places to look.
The recent UN special report on meeting the 1.5°C climate target concluded that rapid, far-reaching and unprecedented transitions were needed across the whole of society, with low energy demand and low material consumption being the top priorities. This seems at odds with using AI to seek more fossil fuels.
But the worst effects will only happen if we let them. My colleagues and I are calling for 20 per cent of all AI research funding to be used to assess its potential benefits and harms, so that we can make informed choices. Technology isn’t destiny. We don’t have to do something just because we can.
This article appeared in print under the headline “Beware an AI-fuelled world”
More on these topics:
Andrew Simms is assistant director of Scientists for Global Responsibility