New text-to-video generator by OpenAI

19 Sept 23:02 Key points summarised by AI

OpenAI unveiled "Sora," a groundbreaking text-to-video AI model capable of generating up to 60 seconds of video from written prompts, still images, or by extending existing footage.
Operating as a sophisticated diffusion model, Sora interprets user prompts to create realistic and complex video scenes, building upon the advancements developed in DALL·E 2 and 3.
While democratizing video creation and making it accessible, Sora also raises significant concerns regarding the potential for deep fake generation, prompting OpenAI to integrate preventative measures against the creation of violent or sexually explicit content.
The technology is currently in its early development phase and is not yet available for public access, with OpenAI acknowledging both its beneficial potential and the risks of misuse.

Macron condemns those who 'sought to bring down the Republic' after rioting mars nationwide protests

Yesterday 12:56 Key points summarised by AI

President Macron condemned widespread rioting and looting during nationwide protests, blaming far-left activists and criminals for the "unrestrained violence" and vowing accountability for those who "sought to bring down the Republic."
Over 2,400 arrests were made as protests against rising living costs and poor school conditions escalated, causing extensive damage to public buildings and businesses in cities like Paris, Limoges (where City Hall burned), Marseille, and Lyon.
The unrest, which organizers hoped would revive the "yellow vest" movement, saw political leaders trade blame, with Macron instructing prefects to protect public buildings and potentially ban future protests to prevent further disorder.