Elaborate Hoaxes in the Age of AI

This week, I’ve seen a lot of over-dramatization of very simple factual events that seem to be fueled by AI in many ways. Now, these aren’t “caused” by AI; I’m not referring to hoaxes that people have used AI specifically to spread, but things that AI has made worse by the ease with which fake information can be made to look very, very real.

Once Upon a Time

Fifteen years ago, this problem existed, but in my opinion, was severely muted. The concept of biased news is not new by any means and has been an issue for as long as the news has existed. There have always been audiences that are more susceptible to believing in these invented stories, scenarios, and scams, and the media has always catered to them, guiding them to the water they wish them to drink.

\ My concern, and reason for this brain dump, is that with AI, these evil parties seem to be able to cast a much wider net than they could before. They can twist real news into something it’s not with fake videos made by AI; they can pump the internet full of AI-generated content that says whatever they want and cites other AI-generated sources, and they can mobilize an army of influencers that spread their filth like wildfire in an instant.

\ I’ve found that recently, a huge chunk of my time when consuming any form of media is spent asking, “Is this real?” I consistently have to find multiple sources and manually scan them, looking for clues that it was AI-generated, a task that is getting harder and harder by the week. The videos are getting more realistic, the content is written better, and the sources I'm used to relying on are less and less trustworthy.

A Man in the Bean

A group of protestors (in this case, trolls) showed up at Chicago’s Bean with claims that there was a man trapped inside. The protestors claimed that they had found evidence that a wealth of life-support systems had been purchased during the making of the Bean statue, and also attempted to make a connection to a potential missing person (a baby, I believe) around the time of its construction. Trolls exist. They always have and always will. That’s not the issue at hand here. The issue is what happened next. This group was clearly trying to be funny, just causing a stir with some radical idea for their own amusement, but the internet used AI to take the country by storm.

\ While scrolling, I started to see dozens of videos with screenshots of these purchase records, x-ray footage of a person floating inside the Bean, and “eyewitness” reports from someone who claimed they could hear knocking or scratching coming from inside the structure. There were also videos of the Bean being constructed, where you could clearly see the equipment being placed inside. Most, if not all, of these were generated by AI and are completely fake. I knew this, being a sensible human, but I had to admit that the quality created compelling evidence. With less common sense, I would have been easily duped.

Aliens

The recent discovery of 3I/ATLAS has been a goldmine for AI generators and the conspiracy-loving masses. From what I could find, which wasn’t much because it seems a lot of this information is being controlled to stop the spread of disinformation, all we know is that a comet is passing through our solar system. This comet looks like a comet and acts like a comet, but is slightly faster and is not orbiting our sun. One scientist ventured a challenge to the “it’s just a comet” consensus to encourage more critical thinking, theorizing that it was, of course, possible for it to be an alien craft. This set the AI-loving conspiracy nuts on fire.

\ My Instagram feed was on absolute fire with fake videos of this comet with lights being emitted from the sides like a ship, with exhaust clearly venting into space, and fake X-ray shots that showed the internal ship structure and beings inside. Countless videos of influencers pretended to be experts on the matter and talked about potential alien invasions. The worst part of all of this was that each one cited different sources and pulled from different video content. It was easy to assume that a “potential alien invasion” was fake; however, I have to admit, the video content they provided was stunningly believable. The “experts” talking were confident and had plenty of “research” to back them up.

\ The most alarming thing about this entire situation is how fast this misinformation is able to spread and how absolutely believable it can make it. I know this isn’t a controlled situation, and we’ve always had irresponsible people running social media accounts to susceptible individuals, but the use of AI in these fields is making the problem more abundant and harder to discern.

It’s not all bad… but it’s pretty bad.

AI is a wonderful tool that has the potential to make our lives easier. I don’t believe that it is ready for constant use yet, despite it being shoved down our throats around every turn. It consistently hallucinates and makes false claims; it slows my work down more than it speeds it up, and has become a barrier to productivity in most situations I’ve tried to use it.

\ However, I can admit it has potential, and there are small automation tasks that I do find small uses for it. That being said, there’s always going to be a heated conversation around ethics and how we should be using AI.

\ I don’t believe there will be any disagreement, however, that the above cases are the wrong way to use AI. The use of AI to generate content to fuel conspiracy theories and spread them to the masses as facts is dangerous and, frankly, terrifying. I believe we’re only seeing the tip of the iceberg, and I worry that we’re headed for a future where we’ll no longer be able to discern the difference between truth and fiction.

---

Photo by Albert Antony on Unsplash