A new social media app called Sora was launched by the creators of ChatGPT on Tuesday, aiming to capture the interest of users currently engrossed in brief videos on platforms like TikTok, YouTube, and Meta-owned Instagram and Facebook. The fresh iPhone application capitalizes on the allure of creating diverse videos, spanning from animated to hyper-realistic styles.
The surge of such videos flooding social media has raised concerns about the prevalence of “AI slop,” which overshadows genuine human creativity and undermines the integrity of information online. Advancements in AI technology have enabled the creation of highly convincing videos, leading to instances where viewers have been misled. From harmless clips like bunnies on trampolines to deceptive deepfakes depicting wildfires, AI-generated videos have managed to deceive social media users.
The official launch video of the Sora app showcases an AI-generated portrayal of OpenAI CEO Sam Altman in various surreal settings, including a psychedelic forest, the moon, and a stadium filled with enthusiastic fans observing rubber duck races. Initially available exclusively on Apple devices in the U.S. and Canada, Sora offers a personalized experience by suggesting new videos based on users’ previous engagements.
Meta recently introduced its own collection of AI-generated short videos through the Meta AI app. Meta CEO Mark Zuckerberg shared a series of AI videos, such as a cartoon rendition of himself, a group of whimsical beings, and a kitten playing with dough. Both Sora and Meta’s Vibes feature personalized content recommendations tailored to users’ preferences.
Professor Jose Marichal, specializing in political science at California Lutheran University, expressed concerns about the inundation of AI-generated content on social media platforms. He highlighted the prevalence of captivating yet potentially misleading videos, emphasizing the importance of a trustworthy information environment for informed decision-making and democratic processes.
OpenAI addressed some of these apprehensions by emphasizing user well-being in its recent announcement. The company stated its commitment to periodically assessing user satisfaction and offering options to customize feeds, prioritizing posts from acquaintances over strangers.
Sora allows users to incorporate copyrighted content in their videos, a move that may provoke contention within the entertainment industry. Copyright holders, such as film and television studios, have the option to exclude their content from the app’s video feed. OpenAI has engaged in discussions with various copyright holders regarding this policy, with Disney reportedly opting out. Earlier this year, OpenAI advocated for the application of the fair use doctrine to AI training on copyrighted material for national security and competitive reasons.
To safeguard privacy and prevent misuse, OpenAI has implemented measures to prevent the unauthorized creation of videos featuring public figures or other users. Users must undergo a verification process, including a liveness check and approval for the use of their likeness in AI-generated videos.
