ByteDance, TikTok’s father or mother firm, has launched a brand new AI mannequin, the OmniHuman-1. Firms popping out with new AI fashions isn’t too stunning as of late. Nonetheless, the OmniHuman-1 AI mannequin is a video generator. However not like different AI-powered video instruments, OmniHuman-1 doesn’t require written prompts. As an alternative, it makes use of pictures, both within the type of selfies, full-body photographs, and even cartoon drawings, as enter.
AI video and picture era shouldn’t be a brand new idea. Firms like Meta have already got comparable AI instruments of their very own. Nonetheless, what makes OmniHuman-1 notably spectacular is how eerily lifelike these AI-generated movies are. These examples of the OmniHuman-1 AI mannequin in motion present how good the mannequin is. The accuracy of facial expressions, physique actions, and lip-syncing makes it troublesome to differentiate between AI-generated content material and precise footage.
The AI mannequin additionally permits customers to include their audio or video. The animated character within the video can then transfer and communicate in a lifelike method primarily based on the audio or video enter. Nonetheless, this development raises important considerations about deepfakes. As AI-generated movies grow to be indistinguishable from actual ones, the potential for misinformation, fraud, and manipulation grows.
Deepfakes have already been utilized in malicious methods reminiscent of spreading disinformation and cybercrime. The hyper-realistic capabilities of OmniHuman-1 may make issues worse. To that finish, social media platforms must tackle a much bigger duty relating to figuring out AI-generated content material. There must be stronger detection algorithms and clear labels to be sure that customers are usually not misled by these hyper-realistic deepfakes.
