[Flipboard] Man was convinced that the caller was ‘AI’ but he couldn’t prove it - until he asked to do the ‘three finger’ test live on camera

With the advancement of artificial intelligence and deepfakes appearing so real, is there really a way to expose them? Scambaiter Jim Browning (@JimBrowning) a.k.a. Ronnie actually did when he came across a man with a face that made him rethink whether it was actually a real face. Despite showing no apparent glitch, mongo hands, glassy eyes, or physical distortions, the person on his screen seemed fishy, if not bizarre altogether. He needed a trustworthy proof to confirm its human identity, with a simple test. He posted a video on his channel on March 26, which has since been viewed by more than 6.2 million people.

The scammer, as Browning specified in the YouTube description, was from a fake company called “Global Metrix,” a New Delhi-based market research company with branches in Singapore and the UAE. The informer had offered to recover his stolen cryptocurrency. The catch was that Browning knew that it was a scam, so when the company person offered to have a Zoom call, he was a bit surprised. But that wasn’t the evidence that the person on the call wasn’t AI.

On the call, Browning asked the person on the screen to share something that would make him believe that he was not AI. When Browning questioned him, he gave out a sulky laugh and then removed the framed certificate from the wall and flashed it on the screen. Browning’s dialogues still expressed the confusion that he wanted solid proof that the man he was talking to via his screen was real. The man dismissed and instead asked him, “So what’s going on, Ronnie?” Browning cleverly asked him to perform a simple test: Hold three fingers in front of your face. Unsurprisingly, the man exclaimed, “Oh come on, that’s too much!”

The man kept on diverting Ronnie’s requests by first holding two fingers on the screen and then three fingers around his collar, never bringing three fingers in front of his face, exactly what the test required him to do. “That’s too much to ask somebody,” he replied again. For a few moments, the conversation continued in a loop, with Browning urging him to display three fingers in front of his face and the man responding that it’s too much. After persistent requests, the man logged out of the call, ultimately proving his scam.

Deepfake fraud is becoming more and more concerning, with Deepfakes now accounting for around 7% of all fraud attempts globally. According to research by Sumsub, deepfake cases have increased by nearly 245% worldwide in just one year. The rise is even sharper in certain countries. The same report found that deepfake incidents in the United States jumped by more than 300%.

People jumped in the comments section to discuss the horrifying possibilities AI has initiated and the uncanny scenarios that the future might reveal as AI continues to advance. “I kinda thought he was using a face filter to smooth out/change his face, so that’s why he was hesitant to do that because it would break the filter, but if it's all AI, that's scary how real it got,” remarked @zdravdat. “We literally can’t trust anything we see on screens anymore,” consternated @Ennsey.

More on Scoop Upworthy

Jon Stewart explains why AI will never benefit the common man: 'They come for our jobs'

World's most advanced AI robot gives 'terrifying answer' when asked if it could design itself

Dakota Johnson reveals why AI should not be used to create films: 'You can't make art with numbers'