Chinese technology company Kuaishou Technology has released a text-to-video (T2V) generator called Kling that would compete with OpenAI's Sora.
In February, OpenAI impressed us with Sora Demo videos that left us all desperately trying to find the “Login” button. Four months later, we're still waiting for Sora to be released, with no word on when which may occur.
Beijing-based Kuaishou develops content-sharing platforms that it says “make content production, distribution and consumption quick and straightforward.” The company's short-video platform, also called Kuaishou, ranks second after TikTok by way of average each day energetic users.
Creating content for its platforms can be much easier if one didn’t should depend on human-generated content. This could possibly be one in every of the explanations for the event of the T2V tool.
Kling turns text prompts into temporally and spatially coherent videos that look great. Kuaishou says Ring can create videos of as much as 2 minutes at a resolution of 1080p and 30 frames per second.
That's a minute longer than what Sora can produce, based on OpenAI. The other big difference is that Kling has been released to the general public, while Sora remains to be under wraps. If you're in China, or have a Chinese cellphone number and a VPN, you’ll be able to apply now to try the app out.
Like Sora, Kling uses a diffusion transformer architecture. It also features powerful 3D face and body reconstruction technology that may use a full-body image as a prompt to generate a video with smooth limb movements.
If you remember the somewhat cringe-worthy video of Will Smith eating spaghetti from the early days of AI-generated videos, then you definitely'll appreciate how amazing this Kling-generated video is.
Sora from OpenAI is crazy.
But KWAI has just Sora-like model called KLING, and persons are crazy about it.
Here are 10 wild examples you shouldn’t miss:
1. A Chinese man sits at a table and eats noodles with chopstickspic.twitter.com/MIV5IP3fyQ
— Angry Tom (@AngryTomtweets) June 6, 2024
Most of the impressive demo videos that feature quite a lot of movement are short clips. The longer videos are more scenic and contain fewer dynamic elements, which can indicate a number of the tool's limitations.
This clip of a changing scene from the attitude of a train window is sort of impressive.
2. Travel by train and have a look at all types of landscapes through the windowpic.twitter.com/WqF9rlJxbh
— Angry Tom (@AngryTomtweets) June 6, 2024
The visual elements that AI has struggled with to this point are things like fingers, teeth, or natural mouth movements. Here's a powerful clip showing how Kling gets these right in a really natural-looking way.
3. A Chinese boy with glasses enjoys a delicious cheeseburger along with his eyes closed in a quick food restaurantpic.twitter.com/ZOCy0n3gTa
— Angry Tom (@AngryTomtweets) June 6, 2024
The beta version of Kling is in some ways a commentary on the East-West approach to AI. While the West debates AI security, data protection and the hazards of disinformation, China is driving development forward, despite US sanctions which can be attempting to slow it down.
While OpenAI tries to make Sora “secure” or politically correct, we can have to show to China for a good T2V tool since there is no such thing as a US-made one.
