One of the most highly-anticipated AI-related products has just arrived: OpenAI's AI video generator Sora launchedon Monday as part of the company's 12 Days of OpenAIevent.
OpenAI has provided sneak peeks at Sora's output in the past. But,Uschi Karnat how different is it at launch? OpenAI has certainly been hard at work to update and improve its AI video generator in preparation for its public launch.
YouTuber Marques Brownlee had a first look at Sora, releasing his video review of the latest OpenAI product hours before OpenAI even officially announced the launch. What did Brownlee think?
According to Brownlee, his Sora testing found that the AI video generator excels at creating landscapes. AI generated overhead, drone-like shots of nature or famous landscapes look just like real-life stock footage. Of course, as Brownlee points out, if you are specifically well-versed in how the surroundings of a landmark look, one might be able to spot the differences. However, there's not too much that looks distinctly AI-generated in these types of Sora-created clips.
SEE ALSO: How to try OpenAI's Sora right nowPerhaps the type of video Sora is best able to create, according to Brownlee, are abstract videos. Background or screensaver type abstract art can be made quite well by Sora even with specific instructions.
Brownlee also found that Sora-generated certain types of animated content, like stop-motion or claymation type animation, look passable at times as the sometimes jerky movements that still plague AI video look like stylistic choices.
SEE ALSO: 7 wild Sora videos blowing up social media after its launchMost surprisingly, Brownlee found that Sora was able to handle very specific animated text visuals. Words often show up as garbled text in other AI image and video generation models. With Sora, Brownlee found that as long as the text was specific, say a few words on title card, Sora was able to generate the visual with correct spelling.
Sora, however, still presents many of the same problems that all AI video generators that came before it have struggled with.
SEE ALSO: OpenAI's Sora is officially hereThe first thing Brownlee mentions is object permanence. Sora has issues with displaying, say, a specific object in an individual's hand throughout the runtime of the video. Sometimes the object will move or just suddenly disappear. Just like with AI text, Sora's AI video suffers from hallucinations.
Which brings Brownlee to Sora's biggest problem: Physics in general. Photorealistic video seems to be quite challenging for Sora because it can't just seem to get movement down right. A person simply walking will start slowing down or speeding up in unnatural ways. Body parts or objects will suddenly warp into something completely different at times as well.
And, while Brownlee did mention those improvements with text, unless you are getting very specific, Sora still garbles the spelling of any sort of background text like you might see on buildings or street signs.
Sora is very much an ongoing work, as OpenAI shared during the launch. While it may offer a step up from other AI video generators, it's clear that there are just some areas where all AI video models are going to find challenging.
Topics Artificial Intelligence OpenAI
Best Bose QuietComfort Ultra earbuds deal: Save $70 at Amazon
The Rag Dolls of Pompeii
Imperfect Voices
Strangers in our Midst
Best iPad deal: Save $70 on 10th Gen Apple iPad
Inside Russia’s War on Dissent
The Ukrainian Appeal
Forty-Four Thoughts for Cecil Taylor
Broncos vs. Bills 2025 livestream: How to watch NFL online
Escape from Dimes Square
The best and biggest news from Xbox's 2019 E3 press conferenceThis street performer dancing to the BBC news jingle is simply gloriousAstronomers find something lurking under the moon, but 'Destiny' fans know what's upJohn Krasinski trolls 'The Office' coBoomerang nude GIFs are the latest sexting trend'Star Wars Jedi: Fallen Order' debuts franchise's cutest droid, BDU.S. women's soccer team is killing it at the World Cup'Game of Thrones' subreddit raises $130K for Emilia Clarke's charityUber Air announces first international city to trial flying taxisBradley Whitford on being the newest 'Handmaid's Tale' baddieUber wants to be the 'Amazon and Google of transportation'Merciful chrome extension replaces photos of Trump with kittensUnited Nations drops Wonder Woman as honorary ambassadorQueen Cersei strikes down Donald Trump with 1 thunderous tweetThere's a deepfake of Zuckerberg on Instagram. Your move, Facebook.Elon Musk says Tesla pickup will look like it 'came out of a sciBruins fan flipping off the Blues is a Big MoodBoomerang nude GIFs are the latest sexting trendAstronomers find something lurking under the moon, but 'Destiny' fans know what's upHow to look at crap on your phone when you don't have signal this summer More than half of batteries could be produced by recycled lithium · TechNode China's Chery reportedly mulls second European factory in the UK · TechNode Baidu to add 3,000 new jobs for fresh graduates, mostly AI Japan gives $2.4 billion in incentives to Toyota and more for battery making · TechNode Zeekr gets approval to test Level 3 autonomous vehicles in China · TechNode BYD to introduce premium SUV, sedan, and more to Japan: report · TechNode LG Display sells its Guangzhou LCD panel plant to TCL for $1.54 billion · TechNode Chinese EV maker Xpeng reportedly turns to hybrids · TechNode Huawei tops Fortune China's first Tech 50 list ahead of BYD and CATL · TechNode Geely, Volvo joint brand launches first all Baidu, Geely joint brand launches second model to compete with Tesla · TechNode Spain hands $146 million to Stellantis’s battery project with CATL · TechNode BYD adds Pakistan to its global EV production footprint · TechNode Japan’s chip Honor secures new round of investment from China Mobile · TechNode Nokia cuts 2,000 jobs in China for cost DJI Air 3S dual Alibaba yields “good results” from three China drafts national law on labeling AI Lidar maker Robosense posts reduced loss, expects to break even by 2026 · TechNode
1.5532s , 8231.3125 kb
Copyright © 2025 Powered by 【Uschi Karnat】,Warmth Information Network
