By Kerry Gallagher
OpenAI, the company that gave us ChatGPT, has launched their own social media app called Sora. Sora is an AI-generated short form video platform. All content is user generated and sharable in a feed. For now access is invite-only, but it will soon be available to everyone. Open AI has put some measures in place to make the launch of their new app “responsible” including:
- All videos made and shared on Sora have visible watermarks so viewers know they are AI-generated from the platform.
- Videos are embedded with a signature that allows reverse search tools to source them to Sora.
- Filters check for certain content in both inputs and outputs, and in both text and audio formats to prevent the creation and posting of explicit, violent, and other inappropriate material.
- So far, attempts to create videos that include the likenesses of President Trump and many celebrities have failed on Sora.
- With a feature called cameo, users can set preferences and even delete videos in which their likeness is featured, even in the posts of others.
- Teen profiles are not recommended to adult users, adult users cannot initiate messages with teens, and mature content is limited in the feeds of teens. Also there are parental control features offered for teen accounts.
Once Sora becomes available to more people, it will put high quality AI video and audio generation abilities into the hands of more people than ever. Some of the concerning aspects of this new frontier of hyper-realism shared by journalists so far include:
- Increasing likelihood of fraudulent and misleading content: “Increasingly realistic videos are more likely to lead to consequences in the real world by exacerbating conflicts, defrauding consumers, swinging elections or framing people for crimes they did not commit, experts said.” New York Times
- Increasing likelihood that real content will be dismissed as AI-generated: “AI-generated video could become indistinguishable from video shot with cameras, undermining trust in footage of the real world. Sora’s combination of improved AI technology and its ability to realistically insert real people into fake clips appears to make such confusion more likely.” Washington Post
- Copyright questions: There are questions about whether Sora was trained on copyrights content such as text from books, movie films, and Netflix shows. This could lead to users creating content that violates copyrights. Washington Post
- Over-personalization of information: “Now I’m getting really, really great videos that reinforce my beliefs, even though they’re false,” said Kristian J. Hammond, a professor who runs the Center for Advancing Safety of Machine Intelligence at Northwestern University. “The whole notion of separated, balkanized realities, we already have, but this just amplifies it.” New York Times
- More bullying opportunities: “A social-video playground full of real people’s faces is bound to produce horrific cases of bullying, slander and other kinds of harm.” Axios
Parents can have a proactive conversation with their teens before Sora becomes publicly available and teens inevitably flock to it. Ask:
- What might a short form video feed made up entirely of AI-generated content include?
- Depending on how people are depicted using the cameo feature, what might be the impact on real life relationships, emotions, and reputations?
- What are some ways users can create AI videos that make a positive difference?
- What would they do if they saw content that was explicit, violent, or upsetting?
Planning for these experiences gives our kids more power over the decisions they will make when they gain access to the future of social media.