Deepfake videos and artificial intelligence could be the rebirth of trust in media
As we become more and more accustomed to seeing deepfake videos made by artificial intelligence, traditional media – shunned in recent times – will regain its former importance, writes Nick Baines
IT IS the 5th of November 2024, the morning of the US Presidential election. As Americans eat their breakfasts, a video whistles around social media. It’s Joe Biden, speaking to camera from the Oval Office. He calmly explains that if elected today, on account of his age, he’ll only be able to serve two years of the term.
The timbre of his voice is immaculate. He has his usual linguistic tropes and speech patterns. There’s a slightly vacant look in his eyes, but that’s not necessarily surprising for the Octogenarian.
Only later that day, after droves of confused voters see the video on Twitter/X, right-wing news outlets, and the cesspits of Parler, Gab, and Truth Social, does it become clear that this was an AI-generated “deepfake”.
This isn’t just some dystopian plot to warn about the perils of AI. It’s a silly but not entirely implausible scenario that could happen next year (the Republican National Committee published something similar a few months ago, for example). Bloomberg recently reported that deepfake technology companies raised nearly $200m from some of the shiniest and most moralising VC firms.
The internet is only going to be increasingly flooded with AI-generated content like this, be it written, video, audio, or image-based. ChatGPT’s creations will still be full of factual inaccuracies unless properly checked by humans (inaccuracies that tech workers call “hallucinations”, seemingly to remove agency and reduce blame). Unsurprisingly, regulators are miles behind. According to the Pew Research Centre, the number of U.S. adults who want the government to restrict false information online has risen from 39 per cent in 2018 to 55 per cent in 2023.
Amidst all this fear, there’s going to be an unexpected winner. Traditional media. That statement might raise an eyebrow or two from journalists who have watched their industry be decimated by technology over the last 15 years. Editors remember all too well how social media made consumers expect journalism to be free, in turn leading to newspaper revenue models built on digital advertising and the relentless search for “clicks”.
But the reason AI might be good news for media publishers is that it will usher in a “trust economy”. In an era of information overload, a premium will be applied to information sources that are known to be credible. Household-name news outlets can provide clarity amidst all the noise.
There are good examples of this already happening in the media industry. The Washington Post’s “Pinocchio” rating system, which assesses the truthfulness of politicians’ statements, is a vital fact-checker. And despite all parts of the political spectrum complaining about BBC bias, the launch of BBC Verify this year shows that truth is at the top of the corporation’s agenda. Ros Atkins is one of the brains behind that scheme; a journalist who rose to fame with 5-10 minute videos explaining big news stories. The popularity of his work seems to debunk the tired narrative that young people can’t pay attention for longer than 30 seconds.
The mainstream media might soon find its place again in modern society. With a focus on trust, journalists can act as navigators through the upcoming AI revolution. It’s always risky to make predictions about a nascent technology, for fear of retrospectively being dubbed a flat-earther. But I feel as though there is cause for optimism amidst all the AI fear-mongering.