Can You Spot the Forgery? This AI Can…

Colin Tedards
|
Aug 1, 2023
|
Bleeding Edge
|
4 min read

Dear Reader,

In October 2009, one asset appreciated in value by 681,718%…

And it happened in an instant.

The asset in question wasn’t a stock or even a cryptocurrency. It was a painting. This painting, to be specific.

Source: Wikimedia Commons

Previously known as Young Girl in Profile in Renaissance Dress, the painting was sold at an auction in 2007 for $22,000. But the owner of the piece felt that the artwork had been misattributed. So, he sought a second opinion.

Art professionals reviewed the work and concluded that the owner was correct. The real creator, they deduced, was none other than Leonardo da Vinci. The value of La Bella Principessa, as it became known, was marked up to around $150 million.

It was a dramatic turn of events. And it goes to show that proving a piece’s origin can mean the difference of hundreds of millions of dollars.

Proving the work to be authentic might be the hardest part. Even today, the authenticity of La Bella Principessa is hotly contested.

Art experts devote an entire career to a specific artist or time period. This makes the world of professional art authentication a very small and exclusive club.

Researchers from the University of Bradford and the University of Nottingham are hoping new AI-powered models can unlock the real truth behind a painting’s origins.

Experts vs. AI

In 1981, art collector George Lester Winward bought the de Brécy Tondo. Like the painting above, for the past 40 years, experts have largely agreed that the painting is a Victorian-era copy of the Sistine Madonna by Raphael.

The de Brécy Tondo alongside Raphael’s 16th-century masterpiece, the Sistine Madonna

Source: The University of Nottingham

The similarities are striking. But skilled con-artists go to incredible lengths to replicate classic works.

For instance, notorious forger Wolfgang Beltracchi teamed up with his wife to forge over 50 artists. He was successful for a time… making an estimated $100 million from fraudulent art.

He was only caught because of a slip-up. He unknowingly used a paint that contained titanium oxide, a compound that wasn’t available until relatively recently.

Art experts have relied on chemical analysis to sniff out frauds like Beltracchi. But advancements in AI are adding an extra tool for uncovering fakes and replicas.

What’s notable about the de Brécy Tondo is that it’s the first example of an AI asserting that a replica is the real thing.

Using facial recognition technology, researchers determined that the face of the Virgin Mary has a 97% similarity to the authentic Sistine Madonna work. The face of the baby Jesus has an 86% probability of being from the same artist.

The study, scheduled to be published later this year, considers a rating above 75% to be identical. The AI model measures more than just similarities seen with the naked eye. It analyzes the brush strokes and patterns that are consistent throughout an artist’s work.

An art expert is limited to what they can take in… going back and forth between the two paintings as their own biases blur the lines between objective and subjective analysis.

AI doesn’t have the same failings. It can digest the totality of both images and overlay the similarities and differences at varying scales.

I don’t know if the AI is right about the de Brécy Tondo’s origins. But I think the bigger question here is far more interesting.

Who Do You Trust?

We’re quickly entering a world where artificial intelligence is “smarter” than any human expert. The question is, will we recognize its genius or dismiss it as faulty code?

We’ve just seen how artificial intelligence has the potential to disrupt the fine art market, even “disagreeing” with a host of human experts. But we can see how this could easily apply to other areas.

In the years ahead, would you trust artificial intelligence to diagnose you at the doctor’s office? Or perhaps we’ll rely on AI to tell us what news is real or fake. Would you feel comfortable letting an AI set the rate of a new loan? Or assess the value of an insurance claim?

And more importantly, what happens when the AI disagrees with the “experts” in these given fields? Who do you trust?

It may offer new ideas that challenge long-standing theories of biology, physics, and chemistry. And it may propose radically new concepts that at face value may seem foolish… but could be wise beyond our comprehension.

Once again, I share this story to make the point that AI is more than just an investment trend. It has the potential to radically reorganize society and upend some long-held beliefs. It could be a very uncomfortable period for some people…

The question for us is, do we fight against it or embrace it?

Regards,

Colin Tedards
Editor, The Bleeding Edge


Want more stories like this one?

The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.