A new program can animate old photos. But there is nothing human about artificial intelligence

by

As far as AI animated images go, the technology behind these Harry Potter-esque photos isn’t particularly complex.

Users are invited to provide old photos of their loved ones, and the program uses deep learning to apply predetermined movements to their facial features. It also compensates for minor moments not in the original photo, such as revealing teeth or the side of a head. Together it creates a, if not a completely natural effect, then a very captivating effect.

Reactions to the Deep Nostalgia images – tears at the sight of a grandmother’s smile, an uncanny sense of connection to a long-dead historical icon – knock on a mysterious emotional wall between us and this kind of rapidly evolving technology.

We rely on perception and emotion

“The appeal here is that visual images are visceral and compelling and we respond to them,” said Hany Farid, associate dean and principal of the School of Information at UC Berkeley. ‘We are visual beings. Seeing your grandmother or Mark Twain come to life has something fascinating about it. ‘

Fascinating – and yes, a bit scary.

Our brains, advanced as they are, prehistorically respond to things that are almost human, but not quite. This is commonly referred to as the Creepy Valley, and a lot of deepfakes and AI-controlled image manipulations set off this old alarm bell. Even MyHeritage addresses this response in their commentary on the program.

“Indeed, the results can be controversial and it is difficult to remain indifferent to this technology,” their FAQ page reads.

When it is a beloved family member who actually occupies that almost-but-not-all of space, the parts of our brain that love and fear face each other, even when we know very well that what we are looking at is not real.

“The way our brain processes images of humans is different from inanimate objects. It uses neural circuits,” says Farid. “For years we have been able to synthesize inanimate objects, and that completely fools the visual system because we have no preconceived ideas about how they move. But when it comes to people, it lags behind. subtle way, we move and recognize these movements. “

“My sense of wonder can be tinged with a sense of dread” said La Marr Jurelle Bruce, a professor at the University of Maryland who shared the animated image of Frederick Douglass and caught the attention of hundreds of thousands of people online.

AI relies on data and rules

Deepfakes, a sophisticated fusion of synthetic audio and images, have been a bone of contention for digital ethicists for years, especially when it comes to issues of altered pornography and fake videos that could threaten national and financial institutions.
For a more positive application, companies are increasingly turning to the technology to create widely customizable ad campaigns. The result, experts say, is that consumers can feel more attached to a brand or product by looking at the specific way it would fit into their life – on a model with similar proportions, in their own language or in a micro targeted ad that speaks to their interests.

These kinds of applications look for a similar kind of human connection as Deep Nostalgia. But the fact is that artificial intelligence has nothing human.

Farid points out that machine learning, which is the engine of more widely available animation technologies such as Deep Nostalgia, is a field within the wider world of artificial intelligence. Machine learning delves into data and finds patterns. While a program may get better with more input, no intelligence or analysis is involved in how it applies these patterns.

There are many applications that benefit greatly from this type of data.

“When you forecast the stock market, you want patterns,” Farid gives as an example. Or making cancer diagnoses. I don’t need to understand at this point why cancer shows up, I just want to know if it is. ‘

When applied to more human pursuits, the lack of, well, intelligence shows.

The smiling faces of our ancestors, while moving, of course, do not hold up when we drop our suspension of disbelief.

AI-generated renderings of human faces, another threat to the security of our online environment, often contain hilarious glitches where a program, unsure of what to do with irregular things like ears or glasses, spews out small monstrosities that are hidden in otherwise convincing faces.
Even extremely sophisticated, true deepfakes like Tom Cruise’s elaborate ones currently being passed down often have minor inconsistencies that make us question our own sense of reality.

However, those inconsistencies will diminish as the technology develops, and Farid says it’s time for companies to take a critical look at the ethical implications of using it.

“The tech industry has done things because they can and not because they should,” he says. “We need to stop building things because they’re cool and start asking these tough questions before it’s too late.”

For example, technology used to be so good that our emotions can transcend our keen senses.

When using Deep Nostalgia, MyHeritage warns users against uploading photos of living people without their permission, and says it will not add audio options in the interest of user safety.

In the future, maybe another program will be able to fill those gaps, and we will be able to see, hear, and talk to those who have been lost long before us. Such technology will pose astonishing challenges to our security and our sense of reality as we know it.

But when it smiles at us through the comforting faces of our closest loved ones, it will be much more difficult to resist.

Source