Posted on 16th February 2024
There's a big buzz among technology accounts and vloggers on social media following the announcement of Sora, a new AI tool from OpenAI that generates videos from text prompts.
OpenAI posted a series of videos to its social channels to announce the new software – like these TikTok videos showing snow falling in Tokyo, woolly mammoths treading through a meadow, a Victoria crowned pigeon and this YouTube video showing a range of videos made with Sora, including a trailer for a movie about a spaceman, a trip back to the Gold Rush, a flock of paper aeroplanes, and a train journey in Japan:
Many of the videos posted by OpenAI, the creators of Chat GPT, show the prompts they used to create the video. The example, below, of the Amalfi Coast in Italy, is generated with the prompt: “a drone camera circles around a beautiful historic church built on a rocky outcropping along the Amalfi Coast, the view showcases historic and magnificent architectural details and tiered pathways and patios, waves are seen crashing against the rocks below as the view overlooks the horizon of the coastal waters and hilly landscapes of the Amalfi Coast Italy, several distant people are seen walking and enjoying vistas on patios of the dramatic ocean views, the warm glow of the afternoon sun creates a magical and romantic feeling to the scene, the view is stunning captured with beautiful photography.”
OpenAI's CEO Sam Altman tweeted some of the results here (5.8m views):
His tweet got replies from video creators including Mr Beast who posted: "Sam plz don't make me homeless".
Sam replied by asking what AI-generated video he'd like to see – generating the video, below, in response to a request to see a monkey playing chess in a park (930k views):
Vloggers have been quick to share examples of the AI-generated videos with their subscribers and offer their own opinions. Marques Brownlee (18.4m subscribers) called it "simultaneously really impressive and frightening":
YouTubers Some Ordinary Gamers (3.6m subscribers) compared the results with one of the first attempts from a year ago, saying "the results are a bit mind-blowing", while Two Minute Papers (1.5m subscribers) say it "might be the craziest thing I've been able to show you in over 800 videos". AI expert Matt Farmer shared an OpenAI Sora video on Facebook, predicting "This will kill Hollywood".
This tweet from @debarghya_das, editing the Sora videos with an AI version of David Attenborough's voice, has had 6m views:
And this video from YouTuber @Blizaine used AI to add a track to the OpenAI Sora Bling Zoo video, creating a music video.
Twitter/X users have responded the new technology with humour, posting a series of memes with imaginary prompts. Examples include:
@jojammss: "This video was generated by Sora, the new model by OpenAI. Prompt: Man in a red shirt takes a bite out of a large burger and gleefully exclaims how good it is"
@testaccountoki: "ok sora went hard on this one 'instagram reel, shot on iphone, following a car driving through american city, meme with funny caption, realistic, portrait mode'"
@WillHobick: "This 60-second video was generated by Sora. That's the new model by OpenAI. The most advanced text-to-video tool created so far. I'll share the videos here. Absolutely insane. 🤯 #openai #sora Prompt: An ocean video that becomes an internet meme that no one can escape"
@ScottieDidIt: "Sora, Prompt: 'a man in a meeting, finally standing up for something for the first time in his life. He’s wearing a hat that only he can pull off - a fedora with a safari flap. Also, he might have dice in his pocket'"
@ayushpranav3: "Sora by OpenAI is just amazing. Here was my prompt: create a video of a girl playing basketball with a football in tennis dress on a golf course"
Sora isn't yet available for people outside OpenAI to use. On openai.com/sora they say they are taking several important safety steps ahead of making Sora available in areas like misinformation, hateful content, and bias.
They wrote: "We’ll be engaging policymakers, educators and artists around the world to understand their concerns and to identify positive use cases for this new technology. Despite extensive research and testing, we cannot predict all of the beneficial ways people will use our technology, nor all the ways people will abuse it. That’s why we believe that learning from real-world use is a critical component of creating and releasing increasingly safe AI systems over time."
