Voices: Can Mark Zuckerberg really be trusted to oversee the AI revolution?

Meta’s vast resources mean that it will inevitably be at the forefront of this exciting, but potentially dangerous, new tech for years to come. That’s more than a little troubling, especially if you’ve read the former Facebook executive Sarah Wynn-Williams’ bestselling tell-all book Careless People: A Cautionary Tale of Power, Greed, and Lost Idealism about her time at the company.

If you only read one business book this year – if you only read one business book in your lifetime – this is the one.

Wynn-Williams hasn’t been shy about including gory allegations about the behaviour of the company’s top staff. Her book is an exposé, and it does that job well. It is also a great, if sometimes horrifying read.

But the important part is what she has to say about the thoughtless way Meta was prepared to use its groundbreaking tech. For example, Wynn-Williams alleges that the company actively targeted teenagers with ads based on their emotional state, including when they were depressed.

Now add AI into that algorithmic mix. The implications are enough to make you shudder. The company’s attempts to court China, and what was offered to the regime there, have also raised eyebrows. Wynn-Williams made the same allegations before US senators, when she was, remember, under oath.

Now let’s put this in context. The world is in the midst of a hell-for-leather AI race. Donald Trump claims the US has already won it. He might be right. Keir Starmer is determined that Britain should at least find a place on the podium, although government communications bods would probably prefer I use phrases like “leading role” or “pioneering”, or some such.

His latest wheeze is bigging it up with influencers (am I alone in hating the word “influencer”?). On Thursday, the prime minister held a half-day session for around a 100 of them in Downing Street about how they can work more closely with the government – and vice versa, given the stratospheric rise of news intake on social media.

AI has ministers drooling, and with good reason. The British state is in a mess, hamstrung by sparse resources and poor leadership, which often seems more interested in pet projects than it is in solving problems. AI could change that.

Justice secretary Shabana Mahmood thinks it will be able to “predict the risk an offender could pose” and inform “decisions to put dangerous prisoners under tighter supervision”, thus cutting prison violence. The struggling probation service could also benefit, with AI pilots showing “a 50 per cent reduction in note-taking time, allowing officers to focus on risk management, monitoring and face-to-face meetings with offenders”.

Over to the NHS, where we are told that an app using AI to provide physio for people with back pain has cut treatment waiting lists by 55 per cent. I don’t know if I particularly want my back problems handled by a bot, but maybe that’s a consequence of my reading too many dystopian books.

We’re already seeing some of the tech’s negative effects through the elimination of junior positions in the tech industry and the City, exacerbating an already weak UK jobs market. But the government isn’t so keen on talking about that.

All this begs the question: is this the sort of tech we want to be controlled by Wynn-Williams’ “careless people”, who are so obsessed with the bottom line that they never stop to think about what they’re doing?

This is not just a criticism of Meta. Britain’s politicians are currently falling over each other to worship at the altar of the tech bros. They think it’s the future. They might be right.

Perhaps this tech fixes the NHS and gets a sclerotic state that is very visibly failing the British people working again. Lives would be improved as a result. The trouble is, if you mix careless people, they could just as easily combine to create a toxic brew.