On 15 February 2024, OpenAI previewed Sora, a model that creates videos of up to a minute from a written prompt. The sample clips, showing detailed scenes and smooth camera movement, were far more convincing than earlier attempts at AI video. At the time, Sora was shared with a limited group of testers rather than the public.
How it relates to image generators
Text-to-video builds on the same ideas as AI image tools, but it must also keep objects, lighting and movement consistent from one frame to the next, which is much harder.
Opportunities
- Quick storyboards and visual ideas for teachers, filmmakers and small businesses.
- Explainer videos that would once have needed a studio.
Concerns
- Fake footage can spread false stories. Seeing is no longer believing.
- Creative jobs and rights. Writers, actors and artists are asking how their work is used and paid for.
- Labelling. Many argue that AI-made media should be clearly marked.
A habit worth teaching
Before sharing a video that shocks or amazes you, ask: who posted it first? Is it reported by trusted outlets? Does anything look odd, such as hands, text or shadows? Media literacy is now part of digital literacy, and children can learn it alongside coding.
