can we really control AI?

This was a big week for artificial Intelligence as the UK hosted the first summit on AI safety at Bletchley Park, where Alan Turing, a godfather to AI, helped crack the German codes during the Second World War. AI safety has been an acute worry for tech and political leaders alike over the past year as the arrival of ChatGPT and other sudden leaps in the technology such as cloning human voices have raised profound existential questions.

If you’ve been following coverage of the summit, you could be forgiven for finding the language unsettling. The Bletchley Declaration, a pledge to work together to regulate AI signed by 26 world leaders, states that AI has the “potential for serious, even catastrophic, harm” and could be used to create biological weapons.

At the end of the conference Elon Musk said that AI would spell the end of work as we know it. “You can have a job if you want a job ,” he said in a talk with UK prime minister Rishi Sunak. “ But AI will be able to do everything.”

But what’s the reality? Is AI out to get us? Can we ever truly regulate it? We spoke to the FT’s artificial intelligence editor Madhumita Murgia, who was at Bletchley Park this week.

She said that the tone of the summit was productive and positive. “I think there have been really diverse opinions. Most people that I spoke to came out feeling quite positive about an exchange of ideas.

“There were lively debates, particularly around open source versus closed source and whether the risks demanded some kind of [safety] stop, and what the plans should be for the next year, as well as for the next five years.”

Madhumita described a spirit of “healthy competition”, especially after the US announced plans to create its own AI regulatory body. “We see it as a race to the top, which is only a good thing,” she said.

AI regulation has been moving quickly in recent months, with another “mini” summit scheduled to take place in South Korea in six months, and a fully-fledged summit planned for Paris this time next year. “There clearly is a willingness and a will on the side of all participants to do this as a regular thing,” said Madhumita. “Everyone agrees the UK has pulled off something quite special.”

Our favourite pieces

This week’s piece about the looming global rice crisis was FT journalism at its best. It’s not a story you’ll see covered on the front pages or by more generalist news outlets, but it could well end up affecting us all. A similar crisis in 2008 drove inflation across the world, sparked unrest in Africa, Asia and the Caribbean and, ultimately, contributed to the Arab Spring. What will the consequences be this time? 
Hannah Rock
Deputy editor, FT Edit (@HannahRockFT)

BlackRock released the results of a study this week that companies with and near gender parity perform better than their gender-skewed peers. Brooke Masters writes that the results come at an interesting time as critics rail against ESG investing.
Caryn Wilson
US editor, FT Edit (@CarynAWilson)

Something to listen to

Working It — How does the UK fix its productivity problem? The Working It podcasts looks at the most, and least, productive countries in the developed world.

Rachman Review — The dollar is used all over the world. So is cutting off access to the currency one of the most powerful tools in the US diplomatic toolbox?

FT Weekend — Even if you’re not a fan of Taylor Swift, it’s been difficult to avoid the pop star over over the past year. Her new film Eras chronicling the US leg of her tour. FT Weekend discusses the film, and the cultural tidal wave the pop star has caused.

Something to watch

Over 10 years China has lent close to one trillion dollars through its Belt and Road development fund. The FT’s global China Editor James Kynge takes a look at what benefit China gets from the fund and its mounting debt problem.