OpenAI's Sora text-to-video tool will be accessible to the public later this year, says CTO Mira Murati.
Sora generates lifelike scenes from text inputs.
Audio integration planned for enhanced realism.
Users can customize Sora's video outputs.
Data sources for Sora's training remain undisclosed.
Sora likely uses publicly available or licensed data for training.
Content sourced from Shutterstock via partnership.
Powering Sora is more costly compared to DALL-E.
Concerns about generative AI's potential misinformation grow.
Sora expected to refrain from generating public figure images.
Swipe Up For More Stories