Project Gucciberg offers classic audiobooks read by an AI deepfake of Gucci Mane

Ever wanted to have Leo Tolstoy Anna Karenina Or by Franz Kafka transform Did Manny Ham Mind, the creator of hits like “Lemonade” and “Wasted,” read it? Well, a) it’s a very specific desire, and b) your lucky day.

Project Gucciberg is the latest drop from the virus plant MSCHF. MSCHF used machine learning to create audio deepfakes that read some classic texts by Gucci Mane. Little girl on Beowolf. They are all free to listen to and come with a book cover that blends perfectly with Gucci Mane’s prolific recordings.

that much What The content of Project Gucciberg is incredibly simple, but Why It’s harder to answer. If you are not familiar with MSCHF, we recommend last year’s costume profile. ^^ ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;;;;;;;;. Essentially, they’re a group of VC-funded creators who make weird stuff designed to cause viruses online, such as the squeaky chicken rods and the Air Max 97 sneakers filled with water from the Jordan River, some of which are sold at low cost. And they?? And profit (probably by selling their services to companies that want the goods them I was infected with a virus online).

speaking The Verge, MSCHF’s Dan Greenberg didn’t explain Project Gucciberg’s motives, but he’s happy to talk about the mechanics. Audio deepfakes are now pretty common up to the point where they were used to commit scams (listen to this Joe Rogan’s replica for a good example). To create one, you need a lot of sample data for the target utterance and the correct neural network to learn and copy the mannerisms.

Greenberg says MSCHF has collected about 6 hours of audio from Gucci Mane speaking in podcasts, interviews, and more. Then I made a manuscript of the clip to support the text-to-speech process. To do this, I had to create a “Gucci Pronunciation Key/Dictionary” to better capture the peculiarities of Gucci Mane’s specific argot.

Project Gucciberg’s redesigned book cover is very pleasing to see.

“Gucci’s pronunciation follows a very specific cadence. For example, he uses a much wider variety of vowel sounds than a typical TTS reader,” Greenberg says. “Dictionaries divide words into phonemes (discrete voice gestures) and then the model uses them as building blocks. As a simple example, we need to know which syllables the model needs to remove or whether they will flow from each other through words. We need to know what to say’talm’bout’, not’talking about’, and the Gucci dictionary {T AH1 LMB AW1 T} takes us where there is no’talking about’ text.”

The results are impressive. Although the results are not always completely consistent or provide the best quality, deepfakes certainly sound like humans themselves. “Our fake Gucci Mane often sounds like he’s talking through a bad mic or low-quality internet stream. Part of the reason is that he’s often doing exactly that in his training data,” Greenberg says.

The exact reason Gucci was chosen for this project boiled down to two factors, Greenberg says. One, the rapper has a unique voice, and the second, the Project Gucciberg pun was so delicious that I couldn’t ignore it.

Greenberg added that MSCHF did not ask Gucci to give permission to use his voice. As the site’s disclaimer cunningly pointed out, the whole project raises interesting questions about copyright in the era of AI fakes. “We didn’t write books and we deep faked our voices,” he says. “Is it copyright infringement? Is it identity theft? All training data (records) used to create Project Gucciberg were publicly available on the web. Gucciberg lives in a lovely gray area where everything is new and everything is going on.” It certainly is! The Verge Contacted Gucci Mane through his record label and asked for a response. I’ll update this story when I get a reply.

Is Project Gucciberg more than quick clicks and laughs? Well, not really. But that’s MSCHF’s business and they do it very well. It’s a bit confusing to hear more than a few minutes of the resulting audio, but Greenberg suggests that the world of on-demand deepfake celebrity audiobooks may have its own unique benefits.

“Sometimes… the extreme casualness of Gucci Mane’s narration really makes the text a new light,” he said about the benefits of listening to Kafka’s Deepfake version. transform. “Gregor Samsa comes across as another man who doesn’t really want to get out of bed. You know?”

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Discover

Sponsor

Latest

OpenAI’s state-of-the-art machine vision AI is fooled by handwritten notes

Researchers at the machine learning lab OpenAI have discovered that state-of-the-art computer vision systems can be tricked into using less sophisticated...

The Ultimate 5G Phone: Realme Narzo 30 Pro Review – Top10Brands.online

The Realme Narzo 30 Pro is Realme's latest 5G phone and it's a beauty. This phone has everything you need for a...

Here’s where you can buy a Nintendo Switch

Like everything else that can light up the mood during this incredibly tough time, the Nintendo Switch is out of supply....

The most powerful Wear OS watches are held back by Wear OS

We tested Mobvoi's TicWatch Pro 3 and Fossil Gen 5 LTE, two of the most powerful Wear OS watches available for...

What is a coding boot camp-the advantages of a coding boot camp

Coding boot camps have emerged over the past decade to fill the talent gap. Technology companies had more vacancies than eligible candidates,...