David Guetta uses ChatGPT and uberduck.ai to deepfake Eminem rap for DJ set

I am not sure what gave you the impression that he didn’t call. I don’t think there’s been confirmation either way. I know people aren’t always familiar with Guetta or EDM, but it’s important to know how prolific and genre blending he and his music is. I would give him the benefit of the doubt. He’s at the top of the list of artists who would do this intentionally, but respectfully to cause a discussion. Guetta has many song collaborations with mainstream artists in R&B and Hip Hop. His full songwriting credits list is huge but just looking at the credits page for who he’s done music videos with should give you a clear understanding of him. He needs to show respect to the artist, but he doesn’t need to defend that he did publicly.

https://m.imdb.com/name/nm1615892/fullcredits

From The Pile: An 800GB Dataset of Diverse Text for Language Modeling:

https://arxiv.org/pdf/2101.00027.pdf

7.1 Legality of Content

While the machine learning community has begun to discuss the issue of the legality of training models on copyright data, there is little acknowledgment of the fact that the processing and distribution of data owned by others may also be a violation of copyright law. As a step in that direction, we discuss the reasons we believe that our use of copyright data is in compliance with US copyright law.

Under pre (1984) (and affirmed in subsequent rulings such as aff (2013); Google (2015)), non-commercial, not-for-profit use of copyright media is preemptively fair use. Additionally, our use is transformative, in the sense that the original form of the data is ineffective for our purposes and our form of the data is ineffective for the purposes of the original documents. Although we use the full text of copyright works, this is not necessarily disqualifying when the full work is necessary (ful, 2003). In our case, the long-term dependencies in natural language require that the full text be used in order to produce the best results (Dai et al., 2019; Rae et al., 2019; Henighan et al., 2020; Liu et al., 2018).

Copyright law varies by country, and there may be additional restrictions on some of these works in particular jurisdictions. To enable easier compliance with local laws, the Pile reproduction code is available and can be used to exclude certain components of the Pile which are inappropriate for the user. Unfortunately, we do not have the metadata necessary to determine exactly which texts are copyrighted, and so this can only be undertaken at the component level. Thus, this should be be taken to be a heuristic rather than a precise determination.

1984. Sony corp. of america v. universal city studios, inc. 2003. Kelly v. arriba soft corp. 2013. Righthaven llc v. hoehn.

The rules for live DJ sets is very different from copyright rules that limit published music. This is the same as when artists use samples of meme audio, which otherwise would be copyrighted. Guetta has a long very large body of work and deserves the benefit of the doubt when it comes to the question of Eminem okaying it. He very clearly wants there to be a discussion about this. Celebrity impersonators for singing have been a thing for a long time. This isn’t about one artist in particular.

I was playing with Tortoise TTS and was genuinely surprised by how good it is with just a few minutes of clean audio. It didn’t take me hours to train or fine tune, the generation step is sort of long, 5-7 minutes for 30 seconds, but it feels really similar to stable diffusion where you do quick test with slow samples and iterations to find a decent seed, and then you let it do a more complete regeneration. It’s zero shot generation that ran on my laptop 2070 max q and i7 10750h. It’s not perfect but it’s believable when layered with music.

https://github.com/152334H/tortoise-tts-fast

Hardly an explanation. I really wanted to understand.

Elevenlabs, by some ex-googlers: https://beta.elevenlabs.io/ If anything, the results I've heard from this are better than the demos.

And VALL-E by Microsoft, which isn't quite as good, but notable because it can clone real voices with only three seconds of training data: https://valle-demo.github.io/