Eyewitnesses, like all the rest of us, don’t see the world around us merely as it is. Our ideas about the world influence how we interpret everything we see (e.g., Sharps, 2024).

Take Christopher Columbus, for example. Like many seamen, he knew stories of mermaids; but unlike many seamen, Columbus actually reported seeing them, although he said they weren’t nearly as attractive as he expected (e.g., Bergreen, 2011).

He was actually looking at West Indian manatees; but he’d never seen manatees before, so he interpreted them in terms of his prior frameworks for understanding (e.g., Bransford & Johnson, 1972). Those frameworks included mermaids, but lacked manatees.

The same psychological factors still influence us today. There is a wonderful ancient bas-relief, held today in the British Museum, depicting what are frequently characterized as “ancient divers.” It includes images of men swimming with their mouths pressed into what appear to be inflated goatskins.

The basic concept on many internet sites, and on a variety of “documentaries,” is that this relief depicts ancient people diving with, essentially, an antique form of SCUBA apparatus. People really believe this.

However, this bas-relief, about 2900 years old and hailing from the ancient Assyrian city of Nimrud, doesn’t depict any such thing. It actually shows soldiers of Assyrian King Ashurnasirpal crossing a river. If you look closely at the relief, you will also see swimming horses and troops ferrying chunks of chariots across the same river.

But what about the men breathing away into their inflatable goats, perhaps, according to the internet and satellite TV programs, preceding Jacques Cousteau’s actual invention of the Aqualung by many centuries?

Well, the fact is you can’t actually SCUBA dive with an inflated goat skin. The air regulation would literally be murder, and any necessary goatskin-related breath-holding, with a rise of more than 3 or 4 feet in depth, could cause permanent lung damage (e.g., Griffiths, 1985). Any poor ancient Assyrian who attempted goat-assisted undersea exploration would almost certainly be literally dead in the water.

So, what are these goat-breathing Assyrians actually up to? For many centuries in parts of Asia, goatskins have been sewn up to provide flotation devices, a somewhat meat-based version of water wings. The soldiers in the ancient relief are actually depicted using the goatskins to improve buoyancy; and the reason these damp Assyrians are depicted as breathing feverishly into their goatskins is that when you make a bag out of a piece of a goat, it’s not exactly airproof, and you have to keep blowing into the thing like crazy to keep it reasonably well-inflated.

Now, this is hardly a revelation, and it’s been known for decades by Western archaeologists and historians. Even in the 20th century, Lowell Thomas (1928) provided first-hand accounts of this method of crossing rivers. Yet this obvious, correct interpretation of the relief is repeatedly discarded by internet aficionados and “documentary” producers. Their prior mental frameworks focus on the modern world of SCUBA divers, rather than on the unfamiliar world of using a goatskin to cross a river. We human beings often interpret pictorial evidence with reference to current ideas, rather than to the actual provenance of the given picture.

The Role of AI in Shaping Perceptions

This is where AI comes in, and not necessarily to our advantage. AI allows its creators to manipulate the frameworks within which apparent depictions of reality are interpreted, when in fact the relevant depictions may never have had any basis in reality at all.

There is currently a widespread internet image purporting to depict members of a specific political group, who are said to be part of an effort to resist specific federal U.S. authorities. Politics aside, it’s a lovely picture, and it looks absolutely real. Granted, the weapon held by one of the protagonists would be a very bizarre choice for an insurrection, but as we’ve seen in previous posts of the Forensic View and elsewhere (Sharps, 2024), weapon selection, from a forensic standpoint, is frequently more an issue of psychology than of tactics.

This AI image looks pretty convincing; it looks like a piece of reality, capable of influencing our frameworks for understanding. However, there are a few problems.

If you expand the image, you find that one of the protagonists’ hands is on backwards. You will also find that another protagonist has only three (very thick) fingers. Finally, the building behind the revolutionary group, although it appears quite official and it does have a beautiful tower attached to it, is adorned with the inscription ”CITY HALL TOWER.”

Now, if there’s a city in the world that has “CITY HALL” inscribed on its city hall, I’ve never seen it. Nor do towers typically bear the inscription “TOWER”; you can usually tell they’re towers without that.

The problem is that the tower, and the building, and the armed gathering in front of it are complete fakes, the renderings of AI rather than of reality.

AI, like any technology, is subject to improvements. Many AI images of today are demonstrably superior to those of even a year or so ago, and they will continue to improve until nobody actually writes “CITY HALL TOWER” on a city hall tower anymore.

AI depictions will become increasingly believable, and witnesses and viewers will be correspondingly affected by the enhanced quality of the relevant AI productions.

Courtroom proceedings and criminal investigations of the near future are therefore likely to become vastly more expensive than they are at present. It will be necessary for video experts to sift reality from fantasy in any given image presented in court, and in any image important for a criminal investigation. It will be necessary to understand fully, on a psychological basis, the foundation of what any given eyewitness believes; was that foundation based on reality, or on AI?

These are significant challenges; but we must be aware of them, and act on them proactively, if we are to make any sense of them in any reasonable criminal justice system.