Taylor Swift gave a performance in Melbourne earlier this year.
Graham Denholm, working with TAS Rights Management, utilizes TAS24 and Getty Images for his projects.
Recent studies suggest that Taylor Swift’s interest in artificial intelligence extends to a hypothetical scenario where AI-generated music replaces all existing recordings, including iconic pieces by historical figures like Ludwig van Beethoven and The Beatles. While this speculative narrative paints the pop sensation as the mastermind behind a musical revolution, experts caution that such a reality is far from certain.
In a joint publication, Nick Collins from Durham University in the UK and Mick Grierson from the University of the Arts London issue a stark warning about the urgent need to address potential conflicts arising from AI advancements before it’s too late.
Fortunately, the likelihood of an AI-induced “Swiftpocalypse” remains slim. Collins clarifies that the concept serves as a theoretical exercise aimed at prompting experts to devise preventative measures against the erasure or manipulation of various forms of data, ranging from artistic creations to scientific knowledge and historical documents.
The researchers envision a future scenario where society heavily relies on a few key data repositories, such as Apple and Spotify. In this context, a malevolent AI could potentially infiltrate these central hubs and tamper with the stored information, posing a significant threat to data integrity. Collins predicts that within a millennium, disputes over the authenticity of audio recordings could arise due to such technological vulnerabilities.
By leveraging current AI capabilities, the researchers demonstrate how AI algorithms can already reinterpret classic songs like Queen’s “Bohemian Rhapsody” or Frank Sinatra’s “I’ve Got You Under My Skin” in a Taylor Swift style. This exercise underscores the potential for AI to manipulate accessible data. However, the extensive transformation of the entire music catalog into “Taylor’s Versions” would come at a substantial energy cost and financial investment, which Swift could feasibly manage.
Collins emphasizes that our reliance on digital and physical storage formats may lull us into complacency regarding data security, leaving us vulnerable to AI-driven breaches. Despite efforts to safeguard societal interests, unforeseen threats could emerge in the future, posing new challenges.
While some experts, like Sandra Wachter from the University of Oxford, acknowledge AI’s capacity to perpetuate biases, they remain skeptical of the doomsday scenarios depicted by Collins and Grierson. Wachter argues that the notion of AI autonomously pursuing its objectives is unrealistic, akin to speculating about alien invasions.
Carissa Véliz, also from the University of Oxford, advocates for proactive measures in AI governance, emphasizing the importance of establishing robust oversight mechanisms rather than resorting to extreme solutions like a “kill switch” for rogue AI models. Véliz warns against overreliance on AI, cautioning that unchecked integration could lead to pervasive biases and misinformation, necessitating careful management to mitigate risks effectively.
Despite the speculative discussions surrounding AI’s potential threats, Taylor Swift’s rumored interest in this domain remains unconfirmed.
