What Freddie Mercury’s Synthesized Voice Sounds Like?

Michael Jackson dancing Gangnam Style, Paris Hilton in an MMA fight, and Winston Churchill performing a striptease act. Even today’s technology allows synthesizing voice and editing videos with modified reality. For the time being, it requires resources of an entire team, but soon, one will be able to accomplish that in a few minutes. And what happens to content then? Copyrights? Who gets all the money

One of the distinct features of the age we live in is chronic tendency of legislators of being late with responding to the novelties that high tech brings into our lives.

At the same time, blaming people who are responsible for keeping the legislation relevant and in line with the reality for their inability to modify obligatory rules and regulations that govern social relations is a stupid thing to do. They try really hard but technology changes our lives so fast and sometimes in very surprising ways that all that is left to do for all of us including employees and managers of regulatory bodies is scratch our heads in wonder.

The code of copyright laws is one of the most vivid examples. The key provisions of copyright legislation have been established a long time ago and stood the test time, and therefore, they are more or less well suited even for such changes as Internet, which no copyright lawyer could have predicted 75 years ago. I’d agree that the first fifteen years looked more like a Wild West for authors; even today it’s easy to find pirate content, but the mere fact that copyright protection laws, albeit to varying degrees, are in effect in the majority of countries is quite notable.

A desirable and popular content is an extremely lucrative thing. You can secure a copyright over texts, music, graphics, images, which very often generate whole lot of money for the owners and more often than not continue to do so many decades after the author’s death. For instance, The Beatles songs and another four million songs of various popular singers and songwriters are owned by Sony/ATV, a US-based record company, rather than by singers and songwriters themselves. Until just recently, the list of content types that could be copyrighted was an established and clear thing. It appears, however, that the situation is about to undergo radical changes soon.

A rapid advancement of neural networks as tools for modifying and creating audio and video content from scratch would make things that were unimaginable just a short while ago not only possible but just a routine task. Technically, you won’t face any tough issues even now if you’d like to synthesize a voice indistinguishable from Freddie Mercury’s recordings or create a video with a performance by Nureyev that had never happened. Yes, these tasks require resources and effort from a team of highly skilled professionals now; nevertheless, this is totally doable and not a science fiction anymore.

In a few years from today similar tools will be available to everyone interested. Any person or organization with a computer will be able to generate any quantity of audio and video content of any imaginable nature. We’ll see Michael Jackson dancing Gangnam Style, Paris Hilton in an MMA fight and Winston Churchill performing a striptease act. More precisely, judging from the actual content on the internet now, with new capabilities users will certainly far outstrip the above examples; while these times are yet to arrive, I’d prefer to save my nerves and time.

Currently, neither voice, nor appearance may be used as copyrighted objects. A video created by a neural network in which Montserrat Caballe sings an AC/DC song with her own (also synthesized) voice is subject to the intellectual property rights, but only if that’s an actual AC/DC song. And if it’s not?

A mass-scale reincarnation of famous and popular dead men and women is clearly ahead of us. We’ll hear new Kurt Cobain songs, while new Queen singles will quickly follow one another; some time after that, video technology will make it possible to release new movies with your favorite actors who are long gone. After a short period of total craze and excitement that any such content will inevitably be accompanied by, the legislators will gather their wits and manage to expand the applicability of intellectual property rights to cover the new area.

And at this very moment, a boom age will dawn for tech companies that will be quick to grasp the new trend and manage to negotiate the deal with all stakeholders, which would allow them to start monetizing the forcibly-brought-back-to-life celebrities like actors, singers, musicians, and painters in a completely legal manner. It appears, however, that the painters will be the first to go: training a neural network on a set of a couple hundred of Vincent van Gogh’s paintings to let it then paint its own art equivalent in style but absolutely genuine will be much easier than recreating an authentic performance by, say, Burt Reynolds.

So, if you are now contemplating where to invest your free cash, I suggest you watch the companies involved in developing and training neural networks to synthesize audio and video. Algorithms that will be worth tens of billions of dollars in a decade only because they are capable of changing the world are now being developed somewhere out there. There are many greedy copyright owners in the world, and therefore, hardly any one of them will miss the chance to make a big buck from resurrecting talented deadmen.

Maybe, that’s your chance to make a killing too?

P.S. United Traders monitors new tech companies and offers investing in them through its platform. As soon as they become available, new investment ideas appear at unitedtraders.com/investments