Midjourney Video Updates + A Deeper Look at Sora

Curious Refuge
23 Feb 202413:21

TLDRThis week's AI news covers advancements in video production tools, with a focus on Sora's capabilities and limitations for filmmakers. We explore the potential of AI in music generation with Sunno's new model, discuss 11 Labs' text-to-sound effects, and touch on the implications of Gemini's 1.5 pro model for the film industry. Additionally, we highlight the integration of AI art generators into social media platforms, the development of Midjourney 6 and rumors of Midjourney 7, and showcase AI films that demonstrate the technology's creative potential.

Takeaways

  • 😀 The Hollywood Professional Association invited the channel to share their vision for democratized filmmaking at their annual Tech Retreat.
  • 🎥 A comparison between Sora and Runway shows Sora's superior realism, but there are concerns about its rendering time and control over generation for cinematic filmmaking.
  • 🚀 Sora's current limitations suggest it might be more suitable for detail shots or establishing shots rather than for the bulk of a film's production.
  • 😂 Hilarious online examples of 'Sora fails' highlight the unpredictability of AI in video production.
  • 🎬 An event in Los Angeles features a feature-length parody of Terminator 2 created by a team of 50 AI artists, including some Curious Refuge students.
  • 🎵 Sunno's AI music generation model has been updated to version 3, offering faster generations, a dedicated instrumental button, and more language support.
  • 📣 The AI filmmaking and advertising course by Curious Refuge is opening enrollment on February 28th.
  • 🔊 11 Labs announced a new text-to-sound effects model, with a demo showcasing its capabilities for generating sound effects for a SORA demo.
  • 🏆 11 Labs was accepted into the Disney accelerator program, highlighting partnerships between innovative AI companies and studios.
  • 🧩 Stability AI is releasing a new model, stable diffusion version 3, with improved quality and more textual control over inputs.
  • 🌐 Google Gemini paused image generations due to politically incorrect results but is testing a new model that can input up to 1 million tokens of information.

Q & A

  • What significant event was mentioned at the beginning of the video script?

    -The significant event mentioned was the Hollywood Professional Association's annual Tech Retreat, where Shelby and the speaker shared their vision for a democratized filmmaking future with over 800 influential people in Hollywood.

  • What is the main difference between Sora and Runway as depicted in the video example?

    -The main difference is that Sora creates a level of realism that Runway does not currently achieve. However, it was later revealed that Sora may not be the ideal cinematic filmmaking tool due to its long rendering times and challenges in achieving consistent character and scene art direction.

  • What is the estimated rendering time for a one-minute clip in Sora?

    -It takes about an hour of rendering time to create a one-minute clip in Sora.

  • What upcoming event in Los Angeles is mentioned in the script, and who is involved?

    -The event mentioned is a feature-length parody of Terminator 2 created by a team of 50 AI artists, including some talented individuals and students from Curious Refuge.

  • What updates were made to the AI music generation model by the team at Sunno?

    -Sunno updated their AI music generation model to version 3, which includes faster generations, a dedicated instrumental button, and more language support, making it more accessible to a wider audience.

  • What is the significance of the new text-to-sound effects model by 11 Labs?

    -The new text-to-sound effects model by 11 Labs allows users to generate sound effects by simply typing in a prompt, which could potentially automate sound effects in films in the future.

  • Why did Google Gemini pause their image generations?

    -Google Gemini paused their image generations due to some politically incorrect results being generated by users.

  • What capability does the Gemini 1.5 Pro model have that could impact the film industry?

    -The Gemini 1.5 Pro model can input up to 1 million tokens of information, which equates to about 700,000 words, 11 hours of audio, or 1 hour of video. This ability to read and dissect information from a video could be used to create AI-generated films in the future.

  • What is the rumor about Mid Journey 7 mentioned in the script?

    -The rumor about Mid Journey 7 is that it will have AI video capabilities, following the high standard set by Open AI's Sora announcement.

  • What did Elon Musk say about integrating Mid Journey into Twitter?

    -Elon Musk said that Twitter is in conversations with Mid Journey to integrate it directly into the platform. If the deal doesn't go through, he still wants to put an AI art generator inside of X, the new name for Twitter.

  • What is the AI film news channel's approach to sharing AI-generated content with its audience?

    -The AI film news channel curates and shares AI-generated films, discusses updates and advancements in AI tools, and provides insights into the potential impact of these tools on the film industry.

Outlines

00:00

🎬 AI in Filmmaking: Sora vs. Runway

This paragraph discusses the advancements in AI tools, particularly in the field of filmmaking. The speaker thanks the Hollywood Professional Association for an invitation to their Tech Retreat, where they shared their vision for democratized filmmaking. The main focus is on Sora, an AI tool that creates realistic visuals, and its comparison with Runway. It is noted that Sora requires significant rendering time, which may not be ideal for interactive filmmaking processes. Additionally, the speaker mentions the challenges in achieving creative control and consistency with Sora, suggesting that traditional image-to-video workflows might continue to dominate AI video production. The paragraph also highlights an event in Los Angeles where AI artists are creating a parody of Terminator 2, and introduces an updated AI music generation model by Sunno, which offers faster generations and more language support.

05:02

🎵 AI Music and Sound Effects Innovations

In this paragraph, the speaker talks about the development of AI in music and sound effects generation. They demonstrate how to use Sunno's updated AI music model, showcasing its ease of use and the quality of the generated music. The speaker also mentions an upcoming text-to-sound effects model from 11 Labs, which will allow users to generate sound effects by simply typing in a prompt. The paragraph includes a congratulatory message to 11 Labs for being selected for the Disney accelerator program, highlighting the potential for AI to enhance storytelling in collaboration with major studios. Additionally, the speaker discusses the capabilities of the new stable diffusion version 3 model by Stability AI, which offers better quality and more textual control. A game is introduced where viewers are challenged to identify which of four complex images were generated by different AI models.

10:04

🤖 AI Film News and Developments

This paragraph covers various AI film news and updates. The speaker starts by discussing Google Gemini's pause in image generation due to politically incorrect results, but also mentions the impressive capabilities of Gemini 1.5 Pro, which can process up to 700,000 words of information. The potential impact of this technology on the film industry is explored, suggesting that AI could be used to generate films in the future. The speaker also mentions a recent issue with the language model Chat BT, which experienced a bug that was quickly fixed, highlighting the ongoing challenges in language model development. Elon Musk's interest in integrating AI art generators into Twitter is noted, along with updates on the development of Mid Journey 6 and rumors about Mid Journey 7. The paragraph concludes with a humorous anecdote about a Will Smith meme and an announcement of AI films of the week, including 'The Pomegranate Spell', 'The File', and 'I Want to Be Happy'.

Mindmap

Keywords

💡AI

AI, or Artificial Intelligence, refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. In the context of the video, AI is central to the discussion of new tools and technologies that are shaping the future of filmmaking and content creation. For instance, the script mentions AI tools like Sora and Runway, which are used for creating realistic video content.

💡Sora

Sora is an AI tool mentioned in the script that creates realistic video content. It is compared with Runway, another tool, to highlight the differences in the level of realism they can achieve. The script points out that while Sora can produce highly realistic results, it may not be ideal for a back-and-forth filmmaking process due to its rendering time and current limitations in control and consistency.

💡Runway

Runway is an AI platform for creating and sharing videos that is contrasted with Sora in the script. It is suggested that Runway does not come close to the level of realism that Sora can achieve, indicating a difference in the capabilities of these AI tools in the realm of video production.

💡Rendering Time

Rendering time refers to the duration it takes for a computer to process and generate a video or image. In the script, it is mentioned that creating a one-minute clip in Sora takes about an hour of rendering time, which is a significant factor to consider when evaluating the practicality of AI tools for filmmaking.

💡Image to Video Workflows

Image to video workflows involve the process of turning still images into video content. The script suggests that these workflows, where creative direction is possible, are likely to continue being a staple in AI video production, as they offer more control over the final output compared to tools like Sora at its current stage.

💡AI Music Generation

AI music generation is the process of using artificial intelligence to create music. The script discusses an update to the AI music generation model by Sunno, which now has faster generations, a dedicated instrumental button, and more language support, making it more accessible and user-friendly.

💡11 Labs

11 Labs is mentioned in the script as the team behind a new text-to-sound effects model. This AI tool is capable of generating sound effects based on text prompts, which is a significant advancement in the field of sound design for films and other media.

💡Disney Accelerator Program

The Disney Accelerator Program is a mentorship program that supports innovative companies in the entertainment industry. In the script, it is mentioned that 11 Labs has been selected for this program, highlighting the recognition of their AI technology in the context of storytelling and media production.

💡Stable Diffusion

Stable Diffusion is an AI model for generating images from text prompts. The script talks about the upcoming version 3 of this model, which promises better quality and more textual control over the inputs, indicating advancements in AI image generation technology.

💡Mid Journey

Mid Journey is an AI platform for creating images and potentially videos, as hinted at in the script. It is mentioned in the context of potential integration with Twitter and the development of new features like character consistency and improved aesthetics in its upcoming versions.

💡Character Consistency

Character consistency refers to the uniformity and continuity in the portrayal of characters across different scenes or episodes in a film or series. The script mentions that Mid Journey 6 will offer character consistency, which is a significant feature for creators looking to produce AI-generated content with a cohesive style.

Highlights

AI news channel discusses the latest developments in AI tools for filmmaking.

Sora's realism in video creation compared to Runway, with a significant difference noted.

Sora's rendering time revealed to be about an hour for a one-minute clip, affecting its suitability for real-time filmmaking.

Concerns about the control over the generated content in Sora and its current limitations.

Sora's potential for detail shots or establishing shots in film production.

AI-generated Terminator 2 parody feature film by a team of 50 AI artists, including Curious Refuge students.

Sunno's AI music generation model update with faster generations and dedicated instrumental button.

Demonstration of creating a UK rap song about chocolate cake using Sunno.

AI Filmmaking and AI Advertising course opening enrollment on February 28th.

11 Labs announces a new text-to-sound effects model with a waitlist available for the public.

Sound effects generated for a SORA demo showcasing the potential of AI in sound design.

11 Labs' selection for the Disney accelerator program, highlighting partnerships between AI companies and studios.

Announcement of the winner of a $500 prize from AOL in a competition related to AI art generation.

Stability AI's upcoming stable diffusion version 3 model with improved quality and textual control.

Comparison of image generation from mid-journey and stable diffusion version 3.

Google Gemini's pause on image generation due to politically incorrect results and tests of their 1.5 pro model.

Potential impact of AI's ability to read and dissect information from videos on the future of film production.

Incident with AI language model ChatBT malfunctioning and the need for developer intervention.

Elon Musk's interest in integrating mid-journey directly into Twitter or X.

Development updates on mid-journey 6 with character consistency and improved aesthetics.

Rumors about mid-journey 7 potentially including AI video capabilities.

Will Smith's parody Sora video and a subsequent AI-generated video of him eating spaghetti.

AI films of the week: 'The Pomegranate Spell', 'The File', and 'I Want to Be Happy'.