Was the Pope REALLY wearing a puffer jacket?

After a viral deep fake photo of Pope Francis in a huge white puffer jacket went viral and fooled thousands, questions have been raised about how easy it is to create AI images that look just like the real thing.

The image shows the Catholic leader, 86, wearing the flashy item over his robes, along with a traditional white skullcap.

The creator, Chicago utility worker Pablo Xavier, 31, used MidJourney AI art technology to create the statue, but said he had no idea it would go viral.

He told Buzz feed: ‘I try to think of ways to make something funny because that’s what I usually try to do.

“It just hit me: I should do the pope.”

The images of the pope were created by Pablo Xavier, 31, a Chicago utility who plays with Midjourney, an AI art tool, while high on mushrooms

The photos are said to have been taken using an AI-generated AI (artificial intelligence) image generator called Midjourney

One person commented on the photo on social media, writing: “The fact that people find a picture of the Pope in a long puffy jacket exciting is a good indication that AI will have a very easy time destroying humanity.”

It comes as a leading deep fake expert predicts that 90 percent of online content will be generated or created using artificial intelligence by 2025.

The technology takes in billions of images from around the internet and identifies patterns between the photo and the accompanying text.

The software then generates the image in the form of a realistic photo or professional painting.

Nina Schick, author of Deep Fakes and the Infocalypse, has predicted an “explosion” of this type of content as we enter “this new age of incredible entertainment” where even dead actors’ faces can be cloned.

Speaking recently on Radio 4 Today about the future of AI in Hollywood and beyond, Ms. Schick said, “I predict that I think 90 percent of online content will be generated or created by AI by 2025.”

Her comments came as viewers described ITVX’s landmark comedy ‘Deep Fake Neighbor Wars’ featuring a line-up of A-listers including Idris Elba, Adele and Harry Kane as ‘worrying’ and ‘creepy’.

The show features Impressionists, with AI ‘deep fake’ technology used to transform them into celebrities such as Kim Kardashian, Idris Elba and Adele.

So in this ever-confusing world, FEMAIL offers you ways to spot a deep fake.

Melting into its own shadow

Technology magazine The edge pointed out that the photo of the pope shows telltale signs of being a fake.

The main sign was what they called “the edge of a spectacle lens” – where the outline of the puffer coat seems to merge into its own shadow.

Bellingcat founder Elliot Higgins shows what it would look like if Donald Trump were arrested or on the run from police in AI

Deep Fake Neighbor Wars is a new TV series from ITV where Greta Thunberg uses deep fake

One of the fake images showed Trump’s wife Melania screaming after his arrest

It added that the outlines in deep fake images often fade into the background.

And it pointed to a number of other details that betrayed the pope’s image as fake.

These include the hand that didn’t quite hold the coffee cup and the crucifix that didn’t have the right angles.

The Verge added that while AI knows the “surface of reality,” it doesn’t know the underlying rules that govern how physical objects interact.

Background

Often the deep fake will focus in the foreground, like a person. But things can get messed up in the background.

Fake images of former President Donald Trump have recently surfaced after he was indicted for allegedly paying hush money to porn star Stormy Daniels.

In the image of President Trump being arrested, the railings in the background are blurry and do not quite match. They also seem to have slightly odd angles.

In a photo of Melania screaming, the people in the background also look less than human.

In another photo of the Pope on a bicycle, the railings also mismatch, while in deep Boris Johnson fakes, the shadows often mismatch.

Scenes of Trump being tackled to the ground by police officers were created using AI technology by Bellingcat journalist Eliot Higgins.

Some fake footage showed Trump being chased down the street by police officers while his wife Melania screamed.

Others showed the former president in prison in an orange jumpsuit.

The sequence of images was created using Midjourney v5 artificial intelligence image generators.

Hands

AI usually struggles with hands. For a long time it didn’t even know how many fingers each hand had.

In the photo of the pope holding the coffee cup, his hands are crushed.

And in the deep fake of the arrest of former President Donald Trump in Washington DC, the hands also look very suspicious.

In a Buzz feed article, Winger-Bearskin, an associate professor of AI at the University of Florida, said AI doesn’t really understand what a “hand” is or how it connects to a human body.

She said AI will often make realistic people, but with nine fingers on their hands.

She said, “It’s just looking at how hands are represented in the images it’s been trained on.”

As police in New York City and Washington D.C. prepared for potential unrest in the event that Donald Trump is indicted over allegations that he paid hush money to Stormy Daniels, deep fake footage emerged that appeared to show the former president being arrested – but the hands are a gift away

‘Deep Fake Neighbor Wars’: Spencer Jones, the show’s co-creator, insists the comedy doesn’t deal with serious topics and makes it clear that the characters aren’t real (Photo: ‘Stormzy’, ‘Harry Kane’)

She added that AI needs to understand what it is like to have a human body, exactly how hands are connected to it and what their limitations are.

To enrol Science focus, Peter Bentley, a computer scientist and author at University College London, said the results were often bizarre.

He said, “Because they just got a lot of samples of things, it tries to fit everything together as best it can.”

Similarly, AI will often produce more teeth in a person’s mouth than is realistic.

Feet are also often a problem.

But a writer in the American genius wonders if AI is making us think it’s not as competent as it really is.

What are Deep Counterfeits?

The technology behind deepfakes was developed in 2014 by Ian Goodfellow, who was the director of machine learning at Apple’s Special Projects Group and a leader in the field.

The word comes from the collaboration of the terms ‘deep learning’ and ‘fake’ and is a form of artificial intelligence.

The system studies a target person in photos and videos, allowing it to capture multiple angles and mimic their behavior and speech patterns.

The technology gained attention during election season, as many feared developers would use it to undermine the reputation of political candidates.

Text

According to the i newspaper, much of AI revolves around images and is not focused on words. Some AI can spell out letters, but many produce incoherent scribbles.

If an image contains a lot of text that is accurate and readable, chances are it’s not an AI image, but a real one.

In a recent deep fake of Prince Harry picking up a McDonald’s bag, the AI ​​produced gibberish when it tried to type the text.

However, AI experts point out that it is possible to add text to a photo afterwards using Photoshop.

Facial inconsistencies

One of the most common signs of deep fakes is that the facial features of the person in the image may be misaligned or blurred.

According to Business todaythis is often around the rims of the eyes, face and mouth.

Zoom in on the photo of the pope and you can see his eyelids merge into his glasses.

The skin between his eyes and ears is also ‘too’ smooth, but other images show crow’s feet.

In another deep fake photo of Boris Johnson, the eyes are also a giveaway. AI doesn’t know that eyes usually have to look in the same direction.

Joints and edges

AI can struggle when things overlap or touch. Often it fails to capture the intricacies of shadows or transitions between bodies and clothes.

But AI companies are making this better and better – and soon many of these problems could be solved.

Related Post