'Funny, but a little disturbing': Dr Tan Cheng Bock on being the subject of deepfake videos

'Funny, but a little disturbing': Dr Tan Cheng Bock on being the subject of deepfake videos
PHOTO: Instagram / @tanchengbock

By now, you might have been tickled (or horrified) by a video making its rounds on local social media: 16 prominent Singaporean political figures seemingly singing in unison to a Japanese pop ballad from a video game series. 

Yes, it’s pretty freaky. That’s because the clip is made using deepfake technology, which video producer Adrian Chan carried out by using machine learning and artificial intelligence to manipulate the politicians’ portraits. 

[embed]https://www.facebook.com/IamVAcer/videos/10158627266604721/[/embed]

It was something that Dr Tan Cheng Bock found amusing but a little disturbing at the same time, he told AsiaOne. 

The 80-year-old head of the Progress Singapore Party (and its chief hypebeast) is one of the politicians picked to lip-sync to Baka Mitai in Chan’s video, alongside local politicians such as Prime Minister Lee Hsien Loong, Leader of Opposition Pritam Singh, and Reform Party chairman Charles Yeo. 

Yesterday (Aug 9), Dr Tan’s social media team shared his stunned reaction to a series of deepfake videos, all featuring his likeness singing different tunes.

“Technology is too advanced,” he wrote in the Instagram post.

“I don’t like it,” he exclaimed in the video while watching a manipulated version of himself singing the theme song from Japanese anime Demon Slayer: Kimetsu no Yaiba.

[embed]https://www.instagram.com/p/CDp_5VhHDNw/[/embed]

He added to AsiaOne: ”I've only just heard about deepfake tech, when I was shown a compilation of videos depicting pictures of me lip-syncing to some random songs, which I was told became popular fairly recently. 

“They look funny, but are a little disturbing nonetheless, especially when the face of the 'singer' gets warped!”

The former MP of Ayer Rajah constituency, however, offered a reminder that technology — though a powerful tool that can be helpful in many ways — must always be used ethically. 

Thanks to the increasing sophistication of artificial intelligence, ultra-realistic deepfake videos have been proliferating online in the last few years, with most of them being pornographic in nature (manipulating celebrity faces onto the body of adult-film actresses) — but there have been some politically-motivated ones as well. 

“If deepfake tech is used to spread falsehoods, it is not ethical, and it is important that viewers be discerning and always check official sources of information,” Dr Tan affirmed. 

ilyas@asiaone.com

This website is best viewed using the latest versions of web browsers.