The popularity of generative AI is rapidly changing how businesses and their agency partners create and develop content ranging from videos to images for paid and organic media. Big Tech firms are accelerating the uptake of gen AI as they use their considerable resources to develop tools that they want businesses to rely on. A case in point: Meta recently unveiled two generative AI tools designed for video asset creation and image editing.
Introducing Emu Video and Emu Edit
The first new product, Emu Video is a text-to-video generation tool that can create short, four-second videos from a caption, photo, or image, along with a description. Emu Edit, on the other hand, is a video editing tool that allows users to easily modify or edit videos using text prompts.
Derived from the technology of the company’s Emu AI research project, Emu Video is intended to help creators generate short video clips based on text prompts. Emu Video can generate high-quality video content from basic text (e.g., “a dog running in a field”) or a photo paired with a description. Additionally, it can animate user-provided images based on a text prompt, surpassing previous achievements and establishing a new state-of-the-art standard.
On the other hand, Emu Edit is a new picture editing tool offering flexibility through detailed instructions. It handles various tasks, including local and global edits, background removal and addition, color and geometry transformations, detection, segmentation, and more. Emu Edit precisely follows instructions, ensuring that pixels unrelated to the specified tasks in the input image remain unchanged.
Both tools are powered by Meta’s generative AI technology and are designed to make it easier for creators to produce high-quality video content. Meta claims that Emu Video is particularly well-suited for creating social media videos and short-form content, while Emu Edit is ideal for editing longer-form videos and making more complex edits.
Meta has emphasized that while these technologies are not intended to replace professional artists and animators, tools like Emu Video and Emu Edit can empower individuals to express themselves in new ways. Whether it’s an art director conceptualizing a new idea, a creator enhancing their latest reel, or a friend sharing a unique birthday greeting, these technologies offer new avenues for self-expression and creativity.
These projects are slated for future integration into Facebook and Instagram, providing agencies and advertisers with the capability to quickly generate videos and edit in-stream images using text prompts. Meta believes that Emu Video and Emu Edit will significantly reduce the turnaround time between creative and media teams. This will allow for more efficient resource allocation towards strategic tasks like audience analysis, targeting, and campaign optimization.
With the introduction of Emu Video and Emu Edit, Meta seeks to solidifying its position as a leader in AI-powered video creation by making it easier for creators with all levels of experience to make videos.
Reactions to Emu Video and Emu Edit
Both Emu Video and Emu Edit have attracted considerable interest because they simplify tasks that have to date been complicated and time consuming.
As Search Engine Land noted, “Instantly generating various video assets and image edits reduces the back-and-forth time between creative and media teams, freeing up resources for strategic tasks like audience analysis. Quick access to diverse creatives can also serve as inspiration for marketers.”
Venture Beat noted that Emu Video and Emu Edit may “revolutionize text-to-video generation and image editing.”
Tech Crunch indicated that the tools bring us “a step closer to AI generated movies.” However, Tech Crunch also flagged some of the weaknesses of the tools. For instance, Emu Video doesn’t appear to have a strong grasp of action verbs, and the tool is capable of committing inexplicable visual errors, such as toes that curl behind feet and legs that blend into each other. Tech Crunch also warned that Emu Video and Edit are a threat to visual content creators, despite Meta’s claims to the contrary.
What Businesses Should Do
At True Interactive, we believe that businesses should embrace generative AI tools but with eyes wide open. Here are some tips:
- Understand the capabilities and limitations. Familiarize yourself with what these tools can and cannot do (as noted in the Tech Crunch article cited above). Be mindful of their mistakes and their strengths. This knowledge will help you set realistic expectations and plan projects accordingly.
- Prioritize ethical usage. Ensure the content generated adheres to ethical standards. This includes avoiding misinformation, respecting privacy rights, and ensuring the content is not offensive or harmful.
- Maintain brand consistency. Avoid the temptation of creating visual images “just because you can.” AI tools should be used to enhance your brand message, not detract from it. Make sure the content aligns with your brand’s voice, style, and ethos.
- Emphasize creativity and originality. Use these tools to create unique, engaging content. Avoid overly generic or repetitive outputs that don’t add value to your audience.
- Be mindful of legal considerations. Be aware of copyright and intellectual property laws. Ensure that any generated content doesn’t infringe on others’ rights. Tech Crunch questioned whether Meta has honored copyright in sourcing images for Emu Video and Edit, and Meta responded by saying the company uses only licensed content. Meta should make these important points clearer. If you are unsure, do press Meta for clarity.
- Stay updated on technological advancements. AI technology is rapidly evolving. Keep abreast of the latest developments to leverage new features and capabilities. For Emu Video and Emu Edit, this means following blogs (like ours) and digging deeper into Meta’s own content to understand how to use these tools. (Meta’s own blog contains a white paper and how-to detail about Emu Video and Emu Edit.)
- Keep people involved. This is the most important tip. All of the above tips cannot be managed effectively without humans being in the loop. Always have a human manage the entire content creation and editing process to ensure quality and appropriateness. In addition, people are needed to monitor performance of all content (whether AI-generated or not) and use feedback to refine strategies and improve future content. People also needed for broader, more strategic needs such as creating policies and governance models for using AI, an example being how transparent your company will be about the use of AI in your work.
All images sourced from Meta