The potential of AI can be weighty

On Sunday evening I learnt that American Senator Lindsay Graham had died unexpectedly from a sudden illness. Because it was posted on Facebook by a random profile, I couldn't tell whether or not the news of his passing was real or if it had been posted by someone with a disliking for Graham and his politics. I had to Google the news to double-check if it was true. Around four months ago there were posts circulating on social media of President Trump being rushed to hospital in a critical condition. Again I Googled this 'news' to discover it was fake. But people truly thought it was real.

You see, these days it's hard to tell what's real and what's not real online. The fact is that these days very few people get their news from the television; most don't even own a TV set and use their laptop to entertain themselves. Many years ago my professor admitted to me that he stopped watching the likes of CNN and BBC because he didn't trust their reporting to be accurate and unbiased and instead got most of his news from social media platforms like X and threads. Unfortunately, the irony is that you can't really rely on what you read on social media either. Most of these sites have been overrun by people other than professional journalists and news outlets. This is understandable since these platforms were created primarily to enable everyone to connect with one another. But the downside to them is that everyone now thinks they're journalists just because they have a keyboard or camera and easy access to publishing.

It's actually quite frightening what the general public is capable of doing online now. If you want someone to look good or bad all you need to do is become adept at editing, whether it's of a video, a logo or some text. When we scroll through social media we often come across news items that appear to be posted by well known news networks. But, upon careful examination, in some cases, we often find that the logo does not belong to the real network it pretends to be. It only resembles it and that's the whole point of the fake profile, to pass itself off as the real deal to disseminate false information. But determining the accuracy and integrity of a text-based news post to ensure it's from the actual source is relatively easy because we can visit its official site. Or we can ask AI and ChatGPT to verify its integrity.

However, the real problem arises when we see videos on social media. So many show depict people doing or saying something but most of the time it's not them. Many are AI-generated and often they seem so lifelike that, unless you have the skills to examine every technical aspect of that video, you're likely to think the celebrity actually said or did what the video purports they said. But most of the time, though, they are fake.

Whilst AI can be useful for verifying something, or for teaching us about a subject we know nothing, unfortunately AI also has a negative side.

There is, therefore, another irony. We might think that watching AI-manipulated videos featuring presidents, social media influencers and celebrities is highly entertaining. But we should remember that this manipulation of real footage or photos and, indeed, using real footage and photos to create AI videos of any of us, can be devastating. It could be done by people who know us and who wish us ill in order to destroy our reputation and to make our lives difficult.

This is not an imagined scenario. There are images of ordinary people suddenly becoming famous for allegedly doing terrible things to little kids or to animals. No one really knows for certain how true these videos or images are but everyone will pounce on the subject of the image or video and rush to judgement. No one will double-check the backstory behind these images and videos. Even when someone comes out with an opposing narrative, everyone believes the original. The damage is, therefore, done and the poster has succeeded.