How do we stop AI becoming an 'engine of inequality', and make it a force for good?

In 2003, an Oxford philosophy professor drew up a thought experiment: the ‘Paperclip Maximizer’.

What would happen, professor Nick Bostrom asked, if we programmed a super-intelligent artificial intelligence (AI) to produce as many paperclips as possible?

The results aren’t pretty.

Compelled to pursue its task by any means possible, a self-teaching AI may trigger the apocalypse: after all, “human bodies contain a lot of atoms that could be made into paper clips.”

“The future that the AI would be trying to gear towards would be one in which there were a lot of paper clips but no humans,” Bostrom concluded.

• AI was meant to make the DWP benefits system more efficient. Instead it’s brought bias and hunger
• AI is a tool like the paintbrush – it can be used for good things and terrible things
• This man wasn’t ready to say goodbye to his dead cat. So he resurrected him as an AI robot

Luckily for those of us who prefer our atoms inside our bodies, super-intelligent AI doesn’t yet exist. You’re not about to be pulverised into stationery.  

But the paper’s core question remains. What happens if AI’s objectives aren’t carefully designed to align with human wellbeing? What is at risk?

As the Global AI summit kicks off in Paris, it’s a problem that experts are urging policymakers to prioritise.

“Without worker representation, AI-driven productivity gains risk turning the technology into yet another engine of inequality, further straining our democracies,” the general secretary of the UNI Global Union, Christy Hoffman, warned delegates.

The Big Issue has spoken with experts who voiced similar concerns.

“There are still significant concerns about how AI impacts communities, workers and the environment,” said Tim Davies, the director of research and practice at Connected by Data and a Paris summit attendee.

“But corporate interests are the driving force in these discussions… industry hugely outnumbers civil society at these events, for example.”

Bhargav Srinivasa Desikan, senior research fellow at IPPR, echoed this concern. “The only other industry in the world that lobbies as much as big tech is Big Oil and Big Gas,” he told Big Issue.

The world is currently at a critical juncture for the emerging technology, Desikan added.

“We have choice. What kind of societal direction do we want from AI? 2025 is a really pivotal moment.”

Where are we with AI development in the UK?

All of a sudden, AI is everywhere.

From so-called “AI slop” memes littering social media, to the unsolicited assistance of chirpy search chatbots: to go online in 2025 is to be confronted with the rapidly changing face of this powerful technology.

Between 10 February and 12 February, world leaders and tech bosses have descended on Paris to discuss its implications for jobs, culture and global governance.

Attendees include OpenAI’s CEO Sam Altman and world leaders including US vice president JD Vance.

“We’re expecting the announcement of various initiatives over the summit, initiatives on environmental impacts of AI, and on how we can build public interest AI,” said Davies.

“There will also be a kind of leader’s declaration, a political declaration from the heads of state here – that’s looking far weaker than maybe people might hope for because of the geopolitics right now.”

Last month, president Donald Trump shredded Joe Biden’s AI guardrails, ditching safety disclosure requirements for companies developing ‘high-risk’ tech.

The launch of DeepSeek has also upped the geopolitical stakes of this AI arms race. The Chinese firm’s model is revolutionary, explains Desikan, because it “can perform just as well as American big tech companies using, using a fraction of the physical resources.”

“If someone is able to create very powerful AI for far less cost and make it free, maybe shakes up the game,” he added.

In the UK, prime minister Keir Starmer is eager to capitalise on the opportunities of artificial intelligence; last month he announced plans to ‘mainline’ it into the UK’s ‘veins’. From fixing potholes to diagnosing cancer, the PM envisaged a UK transformed by new tech.

“Our choice… is not whether the AI revolution happens. That is out of our hands,” Starmer said. “We must decide whether we want to get ahead and shape that revolution, or sit back passively and wait for it to shape us.”

“Getting ahead” is not a bad goal, Desikan reflects. But a “light-touch” approach to regulation is not the way to get there.

“We’re worried that such an approach might kind of lead us in a similar situation to social media regulation 30 years ago. Now we just have to deal with the technologies that have been given to us – governments struggle to regulate social media giants.”

“We want direction as opposed to mere acceleration. Invest in the spaces that AI can have a really positive impact – places of public value, like healthcare and public transport, not just business more generally.”  

“Directing” the use of AI will also have positive implications for the labour market. In the “worst case scenario” drawn up by the IPPR, 7.9 million jobs could be lost to AI. In a “best case scenario,” no jobs will be lost.

How does the government ensure the latter scenario? “There should always be a human in the loop,” said Desikan.  

Authorities should provide tax incentives or subsidies to encourage job-augmentation over full displacement, and regulate areas like health to ensure “human responsibility of key issues”.

The Big Issue has previously reported on the dangers of embedding AI in the benefits system without human oversight – with one claimant alleging it brings “bias and hunger”.

There’s also the question of privacy; the government ought to ringfence public sector data for the public good, Davies says.

“When it comes to the sources of data for the next wave of AI, that will be our public sector data,” Davies explains..

“There are decades of public attitudes research, which says the public are not averse the data being used for the public good, but they want to know that it’s fair and it’s being used fairly… there is a risk of the creeping privatisation of our public data.”

These topics will be on the agenda at the Paris summit. The risk, Davies adds, is that they are drowned out by the “hand waving and hype” of vested tech interests.

Way back in 2003, Bostrom – who coined the ‘Paperclip Maximizer’ thought experiment – issued a sobering warning for AI positivists.

“We need to be careful about what we wish for from a superintelligence,” he said “Because we might get it.”

The delegates in Paris would do well to remember it.

Do you have a story to tell or opinions to share about this? Get in touch and tell us more. Big Issue exists to give homeless and marginalised people the opportunity to earn an income. To support our work buy a copy of the magazine or get the app from the App Store or Google Play.

Big Issue is demanding an end to extreme poverty. Will you ask your MP to join us?