The problem with AI “artists” – O’Reilly

Machine Learning


performance reel. Instagram, TikTok, and Facebook accounts. If you have any inquiries, we will contact you separately via email. All the staples of the actor’s website.

However, all of this belongs to the AI ​​”actor” Tilly Norwood.

This work represents one of the new trends in AI: AI “artists” that eerily represent real humans (according to the author, this is the goal). Tilly Norwood author Eline van der Velden said she was focused on making the film a “big star” in the “AI genre”. This distinction has been used to justify the existence of AI-generated artists, arguing that they do not take jobs away from real actors. Director van der Velden has made it clear that Tilly Norwood has been made photorealistic to evoke a reaction, which is what the talent agency is reportedly trying to portray, and it’s working.

And it’s not just Hollywood. Leading producer Timbaland founded his own AI entertainment company and launched his first “artist”, TaTa, with music created by uploading his demos to the platform Suno, reworking them with AI, and then adding lyrics.

However, while the emergence of AI “artists” is technologically impressive, it risks devaluing creativity as a fundamentally human act and, in the process, dehumanizing and making creative labor “sloppy.”

Enhance industry at the expense of creativity

The generative AI boom is deeply tied to the creative industries, with profit-hungry machines monetizing every movie, song, and TV show they can. Of course, this was before AI “artists,” but AI has made the challenge even clearer. One of the motivations behind the 2023 Writers Guild strike was to counter the threat of studios replacing writers with AI.

For leading companies in the industry, hiring AI “artists” means less reliance on human labor, lower costs, and the ability to mass-produce products at higher speeds. And in an industry already known for poor working conditions, treating “non-essential” work humanely has great appeal.

Technological innovation has always posed the risk of eliminating certain jobs, but AI “artists” are a whole new monster in the industry. It’s not just about speeding up a process or a specific task, it’s about removing human labor from the product. This means demand will be even more scarce in an industry where it’s already notoriously difficult to make money as a creative, and that’s not even considering the impact on the art itself.

AI “slop” takeover

The industry has always been dominated by a focus on making money over quality. Netflix and Hallmark don’t make every Christmas romantic comedy with the same plot just because it’s an original story, and studios aren’t endlessly rebooting and remaking based on successful art just because remaking a ’90s movie with a 20-something Hollywood star is visionary. But it still has an audience and ultimately requires creative output and effort.

Now, imagine that instead of all those romantic comedies cluttering Netflix, there were AI-generated movies and TV shows starring the likes of Tilly Norwood, with soundtracks coming from AI-generated voices, lyrics, and direction.

The entire model of generative AI relies on regurgitation and recycling of existing data. Indeed, it’s a technical feat that Suno can generate songs and Sora can convert text into video images. It’s not a creative renaissance. From classroom essays to motivational LinkedIn posts, AI-generated writing is already becoming mainstream, not only ruining your EM dash but producing content that is consistently low-quality and robotic. The “singing” and “acting” of AI “artists” is the next phenomenal quality disruptor, potentially alienating audiences who turn to art to feel connected.

Art has a long tradition of being used as a means of resistance and challenging the status quo. Protest music has become a staple of the culture, and is no stranger to the civil rights and anti-war movements in the United States in the 1960s. It is so powerful that there are attempts by political actors to suppress it and punish artists. Iranian film director Jafar Panahi won the Palme d’Or at the Cannes Film Festival. it was just an accident, He was sentenced to prison in absentia in Iran for making the film, but this is not the first punishment he has received for his films. Would studios like Sony and Warner Bros. release songs and movies like this if they could order marketing-compliant content from bots?

A sign during the writers’ strike famously read, “ChatGPT has no childhood trauma.” AI “artists” may be able to carry out the intentions of their creators to a limited extent, but what value is there from generated creations that have no lived experience or emotion, especially if that is what motivates them to make the art in the first place?

What’s more, generative AI is far from neutral. We end up seeing a lot of stereotypical and harmful material, especially without the input of real artists. The fact that most AI “artists” are depicted as young women with specific physical characteristics is no coincidence. This is a reinforcement of a long-standing trend to “female” virtual assistants, from ELIZA to Siri to Alexa to AI “artists” like Tilly Norwood and Timbaland’s TaTa, reinforcing the trope of relegating women to the role of “helpers” designed to serve the needs of users, a clear manifestation of human bias.

Privacy and plagiarism

To make actors and singers look and sound as human as possible in movies, commercials, and songs, they need to be trained on real-world data. Tilly Norwood creator Van der Welden defended himself by saying he only used licensed data, examined thousands of images for his work, and went through an extensive research process. However, “licensed data” does not automatically make data acquisition ethical. Look at Reddit. Reddit has signed a multi-million dollar deal to allow Google to train AI models on Reddit data. The vast amount of data of Reddit users is not protected and is only being monetized by the organization.

AI expert Ed Newton Rex discussed how generative AI is consistently stealing from artists and suggested taking steps to license the data and train it in the public domain for use in production. There are ways for individual artists to protect their online work. You can add watermarks, opt out of data collection, and take steps to block AI bots. These strategies allow you to retain data, but more They’re safe, but given how vast generative AI is, they’re probably more of a safety net than a solution.

Jennifer King, professor of human-centered artificial intelligence at Stanford University, offers several ways to protect data and personal information more generally, including making the default option for data sharing “opt-out” and legislation that focuses on regulation as well as transparency of AI use, which could face an uphill battle with the Trump administration’s efforts to strip states of AI regulation.

This is the ethical home where AI “artists” live. Consider the faces of the real people involved in the creation of Tilly Norwood. Companies may have allowed their data to be used, but artists whose “data” is their likeness or creativity probably did not (at least not directly). From this perspective, AI “artists” are a form of plagiarism.

basically impairs one’s creativity as a human being

Considering how art has been transformed by technology before the advent of generative AI, one could argue that this is not a cause for concern and is simply the next step in the process of change. But photography, animation, typewriters and all other inventions were not used to justify the onslaught of AI “artists”. Elimination method About human creativity. Photography did not replace painting, even though it related to painters, but was a new art form. There’s a difference between having a new experimental way of doing something and using data (particularly data obtained without consent) extensively to create creations that blur the lines between what is human and what is not. For example, Rebecca Shue, a professor of computer art and animation at Syracuse University who teaches the course “AI in Creative Practice,” argues that artists can incorporate AI into their creative processes. But, as she warns, “Although AI provides useful tools, you should create your own original work rather than using what is generated by AI.”

It is difficult to understand exactly how AI “artists” will contribute to human creativity, which is a fundamental part of our expression and intellectual development. See Paleolithic cave art. Even 30,000 years ago, humans created art without having safe food or shelter. Unlike other industries, art was not born purely for profit.

The arts are already economically undervalued, as evidenced by the lack of funding for schools. Today, kids who want to become writers will be exposed to marketing from generative AI platforms like ChatGPT that use these tools to “write” their stories. The result may resemble a story, but it doesn’t necessarily have the creativity or emotional depth that comes from being human. And more importantly, the child didn’t actually write it. Still, the very fact that this AI-generated story is possible suppresses the industrial need for human artists.

How do I move forward?

The profit-seeking powers that be may support AI “artists,” but the same cannot be said about public opinion. The vast majority of artists and audiences alike are not interested in AI-generated art, much less AI “artists.” The power of public opinion should not be underestimated. The writers’ strike is perhaps the best example.

Thus, collective mobilization will be key in the future when challenging AI “artists” against the interests of studios, record labels, and other creative industry establishments. There have already been victories, such as the 2023 Writers Guild of America strike, which resulted in contracts that stipulated that studios could not use AI as credit writers. And because music, film, and television are filled with stars who often have economic and cultural power, the resistance expressed in the media could benefit from more actionable measures. For example, a prominent production company run by an A-list actor might pledge not to feature AI-generated “artists” in its productions.

Beyond industry and labor, the dismissal that art doesn’t matter unless you’re a “star” could also play an important role in changing the conversation around art. This means funding arts programs in schools and libraries and letting young people know that art is something they can do, and that it can be fun and bring them joy. It’s not necessarily about making money or making a living, it’s about expressing yourself and engaging with the world.

The fundamental risk of AI “artists” is that they will become so commonplace that the pursuit of art will feel pointless, and that much of the art we consume will lose its fundamentally human nature. But human-made art and human artists will never go out of style. To achieve this, it is necessary to fundamentally eliminate the existence of human impulses and human-made art. The challenge is to ensure that artistic creation is not relegated to the margins of life.



Source link