OpenAI’s AI video generation tool, Sora, is officially being released to the public today. The announcement was made in a video by tech reviewer Marques Brownlee, where he stated, “The craziest part of all of this is that this tool, Sora, is going to be available to the public around the time this video publishes.”
Uploaded earlier today, Brownlee’s video provided an in-depth review of Sora, showcasing its features and evaluating its performance. Brownlee’s review highlighted Sora’s capabilities in generating abstract visuals, cartoons, and photorealistic videos, while identifying limitations in object permanence, physics simulation, and video sequence coherence. Brownlee, who had early access to the tool, described it as “both horrifying and inspiring.”
The launch is not without controversy. Speculation arose after a now-deleted help page suggested that Sora might not be accessible in the EU or UK at launch. This exclusion mirrors previous OpenAI rollouts, such as the introduction of Advanced Voice Mode for ChatGPT, which initially bypassed these regions due to regulatory considerations.
Despite these regional limitations, Brownlee’s video highlights an impressive array of features in Sora, including remix tools, storyboard editing, and video resolution customization, all powered by OpenAI’s servers for seamless processing. The tool also allows users to animate static images and memes, though it enforces restrictions on content involving copyrighted material and public figures.
As part of its ethical safeguards, every generated video includes a watermark.However, Brownlee noted in his review that while these safeguards are a step forward, they could potentially be circumvented. “This is the new baseline for AI video generation—both exciting and a lot to process,” he said, emphasizing the significance and implications of Sora’s capabilities. While OpenAI is set to release an official announcement about the model, it will likely do so in the next few hours.
