How artificial intelligence is disrupting the music and film world

The specter of the robocalypse haunts Hollywood. This time, the fear is not fiction projected on the screen, but very real: screenwriters fear that they will lose their jobs to voice robots if scripts are written by artificial intelligence (AI) in the future. A few days ago, 11,000 members of the Writers Guild of America went on strike – for the first time in 15 years. The union is not only demanding better pay for authors, but also a restriction of AI in creative projects.

The concern is not unfounded: in 2016, the short film “Sunspring” was released, the screenplay of which was written by an AI. The project was laughed at at the time, the reviews were only benevolent because the film had an experimental character, but was otherwise quite devoid of logic. The technology was still rudimentary at the time.

Today, there are high-performing language models like ChatGPT that anyone can use to become a screenwriter.

If George Clooney also speaks Chinese, no voice actor is needed anymore

According to a study by Goldman Sachs, 300 million full-time jobs worldwide are at risk from so-called generative AI, which can produce text, images or other content – including many in the creative industries. Screenwriters or writers don’t have to worry about AI taking their work from them overnight. Computers won’t be writing another “Bridgerton” series of novels anytime soon and be on bestseller lists.

But with generative AI, it’s possible to serially write mediocre screenplays and new templates to create low-cost productions that can be used in publishing programs or streaming services. Fear of cheap machine competition has now sparked a revolt in Hollywood.

The AI ​​revolution is affecting not only screenwriters, but also voice actors. At first they will probably become obsolete. There are already voice generators with which texts can be set to music with the desired podcaster or author’s voice; Software companies offer full catalogs of AI characters that can be modularly inserted into ready-made video templates such as news programs or product presentations. The customer only has to write a script, after which an avatar speaks the text in the desired language. Why hire voice actors when the AI ​​can speak in German, Italian, Japanese and Chinese with George Clooney’s timbre?

Alejandro Graue is one of the first creative people to lose his job because of AI. The Argentinian actor and voice actor has worked on several projects, including as a Spanish voice for a YouTube channel with millions of subscribers.

When he called up a video one day, he was shocked to discover that the voice had been synthesized by an AI, he told the Euronews channel. When he contacted the production company, he was told in writing that he was no longer needed as a voice actor.

AI today like Napster then: a huge challenge for the music industry

The music industry is also experiencing disruption. The “Financial Times” compares the current situation to the 1990s, when the rise of Internet file-sharing sites such as Napster shook up the market and threatened record labels’ business model. In those days, all you needed was a computer with an internet connection to download a song. Nowadays, all you need is a computer with an internet connection to create a song and upload it to the internet.

There are countless music generators that create songs according to the user’s wishes: mood, genre, beats – everything can be configured, just like in a professional recording studio. With just a few clicks, the songs can then be uploaded to Spotify or any other streaming service.

A particularly popular tool is Boomy, which has already produced 14 million songs. The entire Spotify catalog consists of 100 million songs. Thousands of new ones are added every day. It is unclear whether the majority of these have already been produced using AI. Because not every song is exposed as quickly as Drake and The Weeknd’s fantastic duet, which recently caused a stir but was canceled again due to copyright infringement.

Also, not everything sounds as virtuoso as David Guetta’s DJ set, in which the Frenchman integrated the unmistakable voice of rapper Eminem – naturally generated by an AI. But songs without a celebrity bonus are also interesting: the Chinese streaming service Tencent Music had 1000 songs “recorded” by an AI. Mass-produced off-the-shelf, which was well received by the public: one track was streamed over 100 million times. Studio? sound engineer? Musician? Can you save everything!

Grimes also wants to monetize AI songs

The AI ​​songs on Spotify and other streaming services compete directly with the human musicians. Simply put, they generate their revenue as follows: the revenue (subscription fee for all users and advertising revenue) is divided by the number of streams. If suddenly a substantial part goes to AI songs, there is less left for the human musicians.

Music labels sound the alarm. Universal sent a letter to all streaming platforms asking them to crack down on AI tracks. Spotify has even removed tens of thousands of songs in recent days. However, not because they were produced by an AI, but because they were heard by an AI. AI track makers used an algorithm that continuously streamed their songs to generate more revenue. At least that makes it clear who consumes all the songs that are produced faster than they are heard.

Canadian musician Grimes has devised a legal business model with AI songs. Elon Musk’s ex-lover is known for welcoming the future with open arms and has made her voice available to everyone online.

Anyone who wants to should use them for new songs. The only condition: half of the proceeds go to Grimes. This way she makes sure she earns well, even though the most popular Grimes song is no longer hers.

(aargauerzeitung.ch)