To my surprise, since the last Media Matters column I wrote in DM168, headlined “Journalism undergoing critical twists in the age of social media and AI”, I received more media requests for interviews than ever before to explain to the public what artificial intelligence is. And that’s what it means when they say your work “gained traction”, AI or no AI.

That’s how it often works – you write something, other media pick it up, and then the next thing you are on radio and television. That’s the role of journalism: Make things accessible, understandable; be a conduit of reliable information to the public. And now the concept of journalism for the “public good” is finally beginning to catch on.

AI is specific (technology-generated intelligence, information and analysis, formulating algorithms), but it’s also broad. It’s everything from the world of apps to creating deepfakes such as the now famous Kate Middleton photograph posing with her three children for Mother’s Day. But this was human manipulation assisted by technology, or even technology use assisted by human intelligence. Or “just” editing.

Four big photo agencies pulled the picture from their archives, but it’s still circulating on the internet, of course. The media itself was fooled. Some say the pic was edited and gosh, everybody is doing it. After all, it’s a legal hobby, creating avatars and whatnot. Others are outraged and feel betrayed.

The crux of the matter is that the feeling of uncertainty in relying on the media leads to a lack of trust. This knot has resulted from the conflation of social media, technology and AI with journalism.

Remember cheque books?

Think about it: Isn’t the internet itself AI? We have been using it for decades now, for example, googling away for information because we can. Some of us once used dial-up modems to send emails via the internet. In the space of just a few decades, people now feel they will die if they don’t have access to Google.

Meanwhile, the Chinese don’t have international Google (they have Chinese-assisted search engines) and citizens are not dying in their millions. But of course, we wouldn’t want the government to restrict information as China does.

It feels like the other day people were writing cheques; now, with electronic banking that uses face recognition for apps on your smartphone, it’s an anachronism.

Face recognition for internet banking is AI. Face recognition is also now used at most large airports together with physical passport checks and fingerprints.

There are endless complaints from honest citizens about visa applications and banks’ verification processes. Phew, the paperwork and bureaucracy! It’s because of the crooks among us, of course. There are still people who manage to get away with fraud and corruption and swindling funds in an age of AI, or indeed, through using AI.

That’s why the honest, good citizens suffer. They also suffer the massive losses incurred by journalism because of the global big tech companies’ dominance in the market, which has affected the media all over the world.

Trolls and doxxing on social media

It was a privilege to be part of the delegation of the South African Editors’ Forum that made a presentation to the Competition Commission at the Media and Digital Platforms Market Inquiry, where big profit-making global giants (Meta, Google, Instagram, TikTok, YouTube) were also expected to present their side of the story.

X’s owner, Elon Musk, is snubbing the inquiry. I spoke for just a few minutes on how X was a big culprit in the issue of female journalists being trolled and cyberbullied. 

Read more in Daily Maverick: Battle for the future of media — Daily Maverick’s submission to the Competition Commission

Trolls are artificially created human-assisted stalkers, to be found mainly on X.

The term “troll” became well known in South Africa after the corrupt Guptas were exposed. The family were known to have “troll farms” in India, which they used as “attack dogs” targeting journalists who were writing about them. In other words, these artificial trolls watched and waited and then attacked their targets on social media.

There are oppositional gazes about the market dominance of these global tech firms: Are they demons or do they do good?

What about ChatGPT?

Everyone is abuzz about ChatGPT, the most recent AI whizz-kid on the block. Society seemed to get excited about this app in quite a non-nuanced way.

The general “policy” seems to be that it is a menace, this ChatGPT, but it’s also useful. I haven’t used it to write an essay, and I doubt I will. I like constructing my own sentences, finding references and coming up with a hopefully strong and original conclusion.

Read more in Daily Maverick: AI shaping up to become the greatest geopolitical weapon in history

ChatGPT cannot make a personalised argument. It can produce a worthy essay, that’s for sure, grammatically perfectly too. I am getting good at spotting the ChatGPT work, and it creates more work for academics if we have to interview a student after a seemingly perfect little essay.

Algorithms

When you were a journalist in the print-only world, you were protected from whether your story was getting one reader or 500,000 readers (the circulation of a newspaper such as the Sunday Times in its heyday).

Now AI, through algorithms, shows how many eyeballs a news story has and how much time was spent reading that story.

Your personal information is collected through AI. Algorithms are used by advertisers. For example, if you once (okay, twice then) bought a warm, handmade pair of boots from Kalk Bay in Cape Town online, the same ads from the same shop will keep popping up all the time. It’s not random.

The Kate picture

I go to X to see what’s trending, and often there’s some interesting information. On X you also find out who just died that you know, or what books have just been launched. It’s often there first before it gets into the mainstream news.

I’m not interested in British royal family news, to put it mildly. But in recent weeks, trending almost every day all over the world, including South Africa, has been the hashtag “where is Kate?”

Initially, I thought it could be another Joshlin Smith case, only to discover it’s about the public wanting to know what happened to Kate Middleton, the wife of Prince William, heir to the throne. She had not been seen in public since Christmas Day.

Then the big boom: A manipulated photograph of Kate with her three children. And people cheered or jeered, depending on which side of the spectrum they fall – a supporter of royalty or not.

After being sent out by hubby William, the photograph was first published by Associated Press. Then questions were raised about it being manipulated and Kate “took responsibility” and said she had merely edited it.

Read more in Daily Maverick: The mythology of artificial intelligence — of gods, magic and the wisdom of the elders

What a mess, but who cares? It looks like a lot of people do, given that this issue continues to trend. It’s trending because AI tells us what is trending.

So, AI is telling us what most people in the world are interested in. Gossip, it would seem. They could also be interested in whether the media can spot fakes, deepfakes, manipulated images and whatnot, and then tell the public.

Some of the advice given regarding fake news in the era of AI is to be sensible. But how difficult is it to be sensible, given that even the mainstream media get caught, such as in the case of the Kate picture? DM

This story first appeared in our weekly Daily Maverick 168 newspaper, which is available countrywide for R29.