AI-Generated Content Obscures Creativity and Threatens Human Ingenuity

Creative professional: ‘It’s an alien thing. It shouldn’t exist.’
AI-Generated Content Obscures Creativity and Threatens Human Ingenuity
A photo shows a frame of a video generated by new artificial intelligence tool "Sora," unveiled by OpenAI, in Paris on Feb. 16, 2024. (Stefano Rellandini/AFP via Getty Images)
Raven Wu
Sean Tseng
4/8/2024
Updated:
4/8/2024
0:00
Commentary

The realms of video production and musical composition are hallmarks of human creativity and intellect, showcasing the dedication and innovation of their creators. Yet, the advent of artificial intelligence poses a threat to this unique human prerogative.

The introduction of AI technologies like Sora, Suno, and Voice Engine heralds a contentious era for those in the creative sectors. Some creative professionals have expressed concerns, labeling these advancements “alien” technology and arguing they ought not to find a foothold in human society.

On Feb. 15, OpenAI unveiled several AI-generated videos through Sora, igniting mixed reactions within the creative community. The initial announcement was soon shadowed by apprehensions about the displacement of human roles by AI.

Further stirring the pot, on March 25, OpenAI released a second batch of Sora-generated content. These videos showcased an advanced capability to render dreamlike sequences with logical coherence, as seen in works by Paul Trillo, an artist, writer, and director celebrated by Rolling Stone and The New Yorker.

Similarly, Nik Kleverov, a creative director at an Emmy-nominated agency known for its work in brand storytelling and motion design, leveraged Sora to produce commercial content. The outputs strikingly mimicked traditional Western advertising techniques, sparking conversations about the potential emergence of “Sorawood” and the future of Hollywood.

OpenAI’s leadership, including CEO Sam Altman and COO Brad Lightcap, have actively promoted Sora to film industry leaders, suggesting its integration into production processes. This move, coupled with the increasing reliance on AI by filmmakers and studios, has sparked job fears among directors, voice actors, and scriptwriters.

The concerns culminated in last year’s strike by Hollywood screenwriters and actors, prompted by the burgeoning use of AI in productions and subsequent job cuts. The strike concluded only after the Writers Guild of America secured an agreement with the Alliance of Motion Picture and Television Producers, highlighting deep-seated unease about job security within the industry.

‘It’s an Alien Thing. It Shouldn’t Exist’

The emergence of AI in music creation isn’t a novelty, but Suno has marked a seismic shift, becoming the music world’s equivalent of ChatGPT. Suno’s ability to produce music has challenged the painstakingly crafted works of traditional music producers.

Suno enables individuals without prior musical experience to generate complete and melodic songs within minutes. This process involves writing lyrics with the help of ChatGPT, registering a free account on Suno, and feeding the lyrics to Suno.

Suno primarily uses the Bark model to generate vocal melodies, while the Chirp model handles music and sound effects. It can generate songs based on user descriptions of the song’s genre, title, and lyrics, autonomously analyzing necessary keywords, rhythms, tones, and styles throughout the production process.

Its versatility extends to generating tracks across various languages, including English, Spanish, French, Portuguese, and Chinese, with styles ranging from rock and jazz to rap.

Suno’s capabilities have sparked widespread discussion on platforms like Reddit, with users sharing their experiences of tricking friends with the AI songs.

Brian Hiatt, a senior writer for Rolling Stone, shared his experience during an interview with KCRW radio in late March. While visiting Suno’s office in Cambridge, Massachusetts, he inputted the phrase “solo acoustic Mississippi Delta blues about a sad AI,” the program generated a song titled “Soul Of The Machine” within 15 seconds.

Initially skeptical that AI could produce emotionally resonant songs, Mr. Hiatt was astonished by the piece. He explained his selection of Mississippi Delta blues was because “it was something so human,” something he believed Suno couldn’t replicate.

“I still get goosebumps listening to it because it’s an alien thing. It shouldn’t exist,” he said.

Critiques of Suno suggest its output still lacks the nuanced emotional depth and authenticity that human composers bring to music. Critics point out the occasional awkward rhythms and chord transitions in AI-generated music, indicating a gap between AI and human creativity.

Lee Ji-Shin, a Japanese electrical engineer, raised concerns about the broader implications of AI in music production, pointing out that although AI-generated songs improve production efficiency and reduce time costs, they could exacerbate unemployment among professionals.

He warned of a future where AI could not only streamline music production but potentially replace the cultural and emotional essence of human-created music, suggesting a profound cultural shift towards AI-dominated creative landscapes.

AI-Driven Misinformation Campaigns

In a notable advancement, OpenAI announced the trial of Voice Engine in late March, a tool capable of creating new speech from a mere 15-second sample of a person’s voice, offering emotionally resonant and lifelike output. The tool can translate a person’s voice into various languages, maintaining a unique accent and vocal style across a diverse tonal range.

Notably, OpenAI showcased the ability to adapt an English-speaking voice to multiple languages, including an illustrative example of translating “Trump-style” English into languages like Chinese and French, among others.

The platform enforces guidelines to prevent misuse, such as impersonation without consent and the requirement for voices to be declared as AI-generated. Additionally, OpenAI has implemented watermarking for audio clips to ensure traceability and monitor usage.

However, the creation of human-like speech through AI is fraught with risks, particularly in the context of misinformation. OpenAI has acknowledged these dangers, particularly in sensitive periods like election years, and is seeking feedback from a broad array of stakeholders in government, media, and civil society to navigate these challenges.

Despite OpenAI’s assurances and preventive measures, concerns about the potential misuse of Voice Engine persist. The technology raises the specter of more sophisticated and indistinguishable misinformation campaigns, exemplified by past incidents of AI-generated content misrepresenting public figures, such as using Biden’s voice to urge voting and disseminating videos and images of Trump saying things he never voiced.

AI’s potential to amplify misinformation and create deepfake videos could fundamentally undermine public trust and integrity of elections. And the sophistication of such technology could significantly enhance the credibility and dissemination of false information.

The debate extends beyond the immediate risks of misinformation to encompass broader existential concerns.

Zhuge Mingyang, an independent writer and contributor to The Epoch Times, framed the rapid evolution of AI as not merely an employment issue but a profound existential threat, suggesting a narrative where AI represents a form of alien influence intent on supplanting human intelligence and creativity.

The writer’s perspective reflects deep-seated anxieties about the role and trajectory of AI in human society, framing the technological leap not just as an advance but as a challenge to the very essence of human identity and creation.

Ellen Wan and Kane Zhang contributed to this report.
Views expressed in this article are opinions of the author and do not necessarily reflect the views of The Epoch Times.