In a world where visual content reigns supreme, generative AI opens a a fascinating new chapter for communication professionals 🤯 This technology allows you to create custom images and videosradically transforming the way brands interact with their audience. For communicators, it's an invitation to rethinking content creationby offering unlimited possibilities to captivate, engage and tell stories in ways previously unimaginable.
In this We Are COM Club workshop, we will discover the most useful tools for communicators, with some use case and the benefits et cons from each one. Let's go! 🚀
A Short History of Generative AI
Image-generating artificial intelligence works by analyzing thousands, or even millions, of existing images to understand and learn the various patterns, textures, and shapes they contain. Based on this data, it is able to create new images that respond to specific instructions (a “prompt”) provided by users, like a textual description of what they want to see. Through a process of iteration and adjustment, the AI refines these creations until it achieves a result that seems realistic or artistically consistent.
In just a few years, image-generating AI has made considerable progress. And today, video-generating AI is following the same path, as demonstrated by the videos generated by OpenAI's "Sora" AI.
Dall-e, OpenAI's image generator
DALL-E is an image generator designed by OpenAIknown for having developed ChatGPTIt allows you to create all kinds of visuals from text. The tool is available for free through Microsoft. Copilot or Bing, or subscription with ChatGPTIts most recent and advanced model is DALL-E3.
Regarding copyright, the images generated by DALL-E 3 can be used by their creators as they wishAccording to OpenAI's statement, to respect the work of the creators, OpenAI indicates that DALL-E 3 will refuse requests for images in the style of a living artist.
3 use cases for Dall-e
- Create an illustration for your blog post (like the one in this report)
- Create (brainstorm!) your next logo
- Create a storyboard for your next corporate video
The strengths and weaknesses of Dall-e
✅ Free with Microsoft Suite
✅ Easy to use
❌ This is not (yet) “Photoshop”: it is impossible to precisely modify a generated or existing image
❌ He has a very recognizable style
MidJourney, the king of photorealistic AI
midjourney is an artificial intelligence system capable of creating images from text entered by the user (“prompt“). Midjourney is accessible by subscription, via the Discord platform.
Midjourney subscribers retain ownership of all images created through the platform, even after their subscription expires, granting them complete freedom of use. However, exceptions include the restriction on enlarging other users' images, which requires the original creator's permission, and the requirement for companies with more than $1 million in annual revenue to opt for a specific plan for business use.
3 Use Cases for MidJourney
- Create a illustration for your blog post
- Create your photos, in alternative to stock photos
- Create the realistic portrait of your personas
The strengths and weaknesses of MidJourney
✅ A very powerful tool with an active and highly creative community
✅ New features are released regularly
❌ This is not (yet) “Photoshop”: it is impossible to precisely modify a generated or existing image
❌ It's not intuitive: the prompts are more difficult to "tame" and the interface (via Discord) may discourage users
Adobe Firefly, the AI that respects copyright
Adobe Firefly is a web application integrated into flagship applications from adobe and to Adobe Stock, as well as a Web application.
Adobe emphasizes that its AI technology, Firefly, was trained on legally acquired content, including those licensed under Adobe Stock and works in the public domain or whose license has expired, while offering compensation for the use of these features in order to to guarantee their legality.
3 use cases for Firefly
- Extend the back-plankn of a visual (switching from portrait to landscape format)
- Remove some objects from an existing photograph
- Recolor your vector objects
The strengths and weaknesses of Firefly
✅ Integration into the Adobe Creative Suite, with the power of the tools
✅ An “ethical” approach to image generation
❌ The price, due to its integration with the Adobe Creative Suite
❌ The generated images can sometimes appear very “synthetic”
Synthesia, avatars at the service of video
Synthesia is an AI-powered video creation platform. This online tool generates interactive videos by bringing to life customizable animated avatars, thanks to a speech synthesis system.
You cannot use stock avatars for any form of paid advertising or television broadcasting; to do so, you must purchase a custom avatar.
3 use cases for Synthesia
- En internal communicationfor regular newsletters or practical videos, on the intranet
- En customer communicationfor videos aimed at users, in the style of FAQs.
- For video modules of training
The strengths and weaknesses of Synthesia
✅ A user-friendly interface that's easy to learn
✅ A wide selection of avatars (and the option to create your own, at an additional cost)
✅ Options to adjust the pronunciation of words (brand name for example)
❌ The budget, which can quickly escalate (per-minute pricing)
❌ The voices may still sound very robotic
Fliki, generative AI for tailor-made videos
flick is an online video generator based on artificial intelligence. This text-to-speech program allows users to convert text into video from textual content such as blog posts, scripts, etc.
The editing rights are held by the user of the content; Fliki does not own any content.
3 use cases for Fliki
- Turning a blog post into a video for social media
- Create a podcast, from a script
- Translating and localizing content
Fliki's strengths and weaknesses
✅ A user-friendly interface that's easy to learn
✅ Options to adjust the pronunciation of words (brand name for example)
❌ The budget, which can quickly escalate (per-minute pricing)
❌ The visual generation isn't the best and the voices can sound very robotic
Sora, OpenAI's video-generating AI
Sora is the AI-based video generator developed by OpenAI. Sora is currently in the final testing phase and should soon be available to the general public. It's not yet known whether it will be available on a paid basis or not, but it should be included in the various plans that accompany ChatGPT.
The replay (in English 🇬🇧) and the workshop presentation materials
Download the presentation 🇫🇷 Download the presentation 🇬🇧And to go further, discover our training course “AI for creating high-impact visuals and videos”
