Mona Lisa is brushing up on her rap skills.
Microsoft put out an artificial intelligence-generated video of Leonardo Da Vinci’s Mona Lisa rapping, and viewers have strong opinions.
The tech giant revealed its new AI technology called VASA-1 which can take a still image and an audio clip and combine them to generate lifelike videos of talking faces.
To demonstrate this, Microsoft released a clip of the Mona Lisa rapping to “Paparazzi” — not the Lady Gaga song, but rather a rap Anne Hathaway wrote and performed on Conan O’Brien’s talk show in 2011.
People had strong reactions in response to the viral video, with one questioning, “Why does this need to exist?
I can’t think of any positives.” One person found the clip funny, saying that it had her “rolling on floor laughing.” “Another day another terrifying ai video 😩,” someone lamented.
If only Da Vinci could witness this,” a user quipped.
“I think this rapping Mona Lisa has swizzled my temporal lobes,” one commented.
Rapping is something Mona Lisa is getting back into.
Viewers’ reactions to a video that Microsoft released featuring Leonardo Da Vinci’s Mona Lisa rapping were largely negative.
A still image and an audio clip can be combined to create realistic videos of talking faces, thanks to a new AI technology that the tech giant unveiled called VASA-1.
In order to illustrate this, Microsoft published a video of the Mona Lisa rapping to “Paparazzi,” which is actually a rap written and performed by Anne Hathaway on Conan O’Brien’s talk show in 2011. It’s not the Lady Gaga song.
The video was released as part of research that was published on April 16. It went viral fast, with one post on X (formerly Twitter) featuring the video receiving 7.1 million views as of Monday morning.
Following the viral video, many expressed strong opinions. One person asked, “I can’t think of any benefits; why does this need to exist? “.
A viewer expressed amusement at the video, stating that it made her “roll on the floor laughing.”. “.
“Yet another awful artificial intelligence video 😩,” a person bemoaned.
“This is simultaneously wild, freaky, and creepy 🤯,” wrote another.
Oh my. “I wish Da Vinci could be here to see this,” a user joked.
One person said, “I think this rapping Mona Lisa has swizzled my temporal lobes.”.
According to Microsoft, the “out-of-distribution generalization” approach demonstrates “the ability to handle audio and picture inputs that are outside of the training distribution.”. “.
For instance, it can process artistic images, audio files of singers, and speech in languages other than English. The training set did not contain any of these kinds of data. “.
The business emphasized the advantages of the new AI technology, including how it will improve educational equity, make learning easier for people who struggle with communication, and provide therapeutic or companionship to those who are in need.
It is not meant to “mislead or deceive,” they cautioned, and there is a chance that it could be abused to impersonate people.
Owing to the various factors surrounding AI technology, Microsoft stated that they do not intend to release any online demos, APIs, products, further implementation details, or related offerings until they are confident the technology will be used appropriately and compliantly with applicable regulations. “.
