New FREE AI Video Generator & Feature Length AI Films!
TLDRThe video introduces Hyper, a new AI video generator developed by former Google Deep Mind alumni, offering text-to-video and image-to-video capabilities. It showcases impressive features like video repainting and a user-friendly interface. Despite a 2-second limitation on HD videos, the platform demonstrates high-quality outputs, with plans to extend video lengths in the future. The video also touches on research into generating full-length AI films and recent drama between Mid Journey and Stability AI, hinting at upcoming advancements in the AI video generation space.
Takeaways
- 🆓 A new AI video generator called Hyper is available for free, developed by Yia Meow and Zuu Wong, former Google Deep Mind employees.
- 💰 Hyper has raised over $19 million, indicating strong financial backing and potential for growth in the AI video generation market.
- 🎥 Hyper offers text-to-video and image-to-video generation, including a unique feature of video repainting, which allows changing the style of a video.
- 🖼️ The platform provides various templates and options for video generation, such as full HD, standard definition, and the ability to extend video length.
- 📈 Hyper's HD video generation is currently limited to around 2 seconds, but a workaround using video editing software like Premiere or DaVinci Resolve is suggested.
- 🔍 The script discusses the potential for AI-generated feature-length movies, with research into long video understanding and AI-generated scripts and visuals.
- 📝 The paper on movie LLM (Language Model) describes a process for generating detailed scripts and corresponding visuals for movies using GPT-4 and text-to-image models.
- 🎬 The style immobilization process in movie LLM ensures consistency in the look of the film, which is a significant advancement for AI-generated content.
- 🤝 Stability, a company mentioned in the script, is working on its own AI models and has a data set that outperforms other models, according to the CEO, Ahad Mod.
- 💬 There is a conflict between Mid Journey and Stability, with allegations of scraping for images and text prompts, leading to a ban on Stability employees using Mid Journey's service.
Q & A
What is the name of the new AI video generator mentioned in the transcript?
-The new AI video generator is called Hyper.
Who are the founders of Hyper and what is their background?
-Hyper is founded by Yia Meow and Zuu Wong, both former Google Deep Mind alums.
How much funding has Hyper raised?
-Hyper has raised over $19 million.
What are the main features of Hyper's video generation capabilities?
-Hyper offers text to video, image to video, and video repainting features.
What is the limitation of the HD video generation in Hyper?
-The HD video generation tends to be around 2 seconds long.
How can users extend the length of their HD video clips generated by Hyper?
-Users can extend the length of their clips by using a nonlinear editor like Premiere or DaVinci Resolve, adjusting the speed and enabling optical flow.
What is the research area that the transcript mentions could lead to feature-length AI-generated movies?
-The research area is called 'movie llm' which leverages GPT 4 and text to image models to generate detailed scripts and corresponding visuals.
What was the cause of the 24-hour outage on Mid Journey?
-The outage was caused by bot-like behavior from paid accounts, which originated from Stability employees scraping for images and text prompts.
What was the outcome of the conflict between Mid Journey and Stability?
-Mid Journey CEO David Holtz banned all Stability employees from using the Mid Journey service, while Stability's Imad Mustafak and Nick S Pierre engaged in a public dispute on Twitter.
What is the current status of Mid Journey's video model and Stability's Stable Diffusion 3?
-Mid Journey is still training their video model, which is reported to be quite good, and Stability is preparing to release Stable Diffusion 3.
Outlines
🌟 Introduction to Hyper AI Video Generator
The video script begins with an introduction to a new, free AI video generator called Hyper, developed by Yia Meow and Zuu Wong, former Google Deep Mind employees. They have raised over $19 million, indicating Hyper's potential in the AI video generation market. Hyper offers text-to-video and image-to-video generation, including a unique feature of video repainting. The interface is user-friendly with options for video generation, and the script provides a demonstration of creating a video using the platform.
🎥 Hyper's Video Generation Capabilities
The script delves into Hyper's capabilities, showcasing its text-to-video and image-to-video features. It highlights the platform's ability to generate high-definition videos, albeit with a current limitation of around 2 seconds per generation. The script also discusses the standard definition model, which allows for longer videos but with less dynamic movement. The user interface is explored, including options for extending video length and adding stylistic elements like the steampunk theme. The script compares the output of Hyper with other AI video generators and provides examples of the quality of videos generated.
📚 Research on AI-Generated Movies and Mid Journey Drama
The script shifts to discuss research on AI-generated movies, specifically the Movie LLM, which leverages GPT 4 and text-to-image models to create detailed scripts and visuals. The process involves generating a movie plot, style immobilization, and video instruction by data generation. The script also touches on the challenges of training models for feature-length movies and the use of the MovieNet dataset. Additionally, the script addresses recent drama between Mid Journey and Stability AI, with allegations of data scraping and subsequent bans, leading to a public exchange on social media.
Mindmap
Keywords
💡AI Video Generator
💡Hyper
💡Text-to-Video
💡Image-to-Video
💡Video Repainting
💡Community Feed
💡Movie LLM
💡Style Immobilization
💡Mid Journey
💡Stable Diffusion
Highlights
A new AI video generator platform called Hyper is introduced, offering text to video generation.
Hyper is developed by Yia Meow and Zuu Wong, former Google Deep Mind alums, and has raised over $19 million.
Hyper's platform allows for image to video generation and video repainting.
The interface of Hyper is user-friendly with a light and dark mode option.
Hyper offers various options for video generation, including full HD, image animation, and standard definition with text prompts.
The standard definition model allows for up to 4 seconds of video generation.
Hyper's HD video generation tends to be around 2 seconds, but there are workarounds to extend the clips.
The community feed showcases the diverse capabilities of Hyper, with outputs approaching Sora quality.
The video discusses the potential for AI-generated feature-length movies using Movie LLM, which leverages GPT-4 and text-to-image models.
Movie LLM uses a style immobilization process to maintain consistency in the generated film's look and characters.
The Movie LLM research paper does not provide video examples but discusses the process of generating detailed scripts and visuals.
Mid Journey CEO David Holtz addresses a 24-hour outage caused by bot-like behavior from paid accounts, allegedly from Stability.
Stability's response on Twitter suggests they have not been scraping and are satisfied with their data set and model performance.
Mid Journey is still training their video model, which is reported to be quite good.
Stability is preparing to release Stable Diffusion 3.
The video creator, Tim, provides a quick hack for extending short video clips using nonlinear editors like Premiere or DaVinci Resolve.
Topaz Video AI is recommended for the best results in extending video clips, despite its high cost.
The video concludes with a summary of the week's events in Creative AI, highlighting the ongoing developments and drama in the industry.