Text to Video: AI Video Generation Overview
Introduction
Text-to-video or AI-generated video options are still in their early stages, and what's currently available may be limited. However, several platforms can be valuable for educators, even if creating a fully AI-generated video lecture or tutorial is not yet possible. In this overview, we'll explore available text-to-video options and some unique alternatives that may have potential in educational settings. All the options presented will likely have some associated costs.
AI Generated Video Resources
Here, I’ve curated a list five AI video platforms that may be useful for creating educational content. This list includes both true “text-to-video” applications and some innovative alternatives. I'll include personal insights for each application, recognizing that my opinion may not apply universally.
1. Synthesia: AI video generation platform allowing creation of AI avatars speaking scripts in 60+ languages. Used by major brands. Offers free trial and paid version with more minutes and features.
Website: Synthesia
Note: Great for instructors with camera anxiety, as the AI avatars are quite good at reading your script.
2. Elai: Users can upload their own images or videos as backgrounds or characters. Offers free version and paid version with more animations and features.
Website: Elai
Note: Specializes in avatar presentations, much like Synthesia. And while the lip-syncing is well done, there is still a distinct “uncanny valley” experience.
3. Descript: AI-powered video editing software that edits video by editing the script.
Website: Descript
Note: A promising option to explore for hands-on editing. While I’ve yet to use it, I plan to dive deep very soon, and am quite excited to do so, based upon what I’ve learned.
4. Runway ML: Creative AI-powered tools for image and video generation, with an intuitive interface.
Website: Runway ML
Note: Limitations on clip length (four seconds, only) and consistency may challenge longer video needs. Still, if you need a clip, I would highly encourage you to take some time to play.
5. Pika Labs: Text-to-video platform available via Discord.
Website: Pika Labs
Note: Limited to four-second clips, similar to Runway ML, but free. And, if you’re a fan of the Discord UI and environment, then there’s no reason why not.
What to Watch For
There are some very interesting and curious things being done right now, even if not yet available to the general public. If you are curious to see what might be around the corner, then I would encourage you to explore these examples listed here.
The Simulation: This is a quite mysterious yet fascinating company working on a project called "The Simulation", with a mission reading, “as audacious as it is intriguing, is to birth a new kind of life: the world's first genuinely intelligent AI virtual beings.” Meanwhile, they have wowed the tech-world with an entire episode of South Park created by AI. It’s quite fascinating and quite alarming.
Explore more: The Simulation
Read about the South Park simulation: “AI put me in a ‘South Park’ episode”, Devindra Hardawar, engaget.
Microsoft: With Microsoft teaming up with OpenAI, and recently with Meta, I think it is only a matter of time before Microsoft releasing an image-to-text application available to the general public. Meanwhile, I would suggest you explore what one might do with a bit of skill and time within Microsoft’s Azure OpenAI platform.
Take a deep dive: “Transforming Text to Video: Harnessing the Power of Azure Open AI and Cognitive Services with Python”, Sabyasachi Samaddar
Conclusion
The text-to-video field is evolving rapidly. With the continuous advancements, it's highly likely that we'll soon have fully functional text-to-video platforms tailored to specific educational needs. Stay engaged and explore these tools, as they may revolutionize the way we create and consume educational content.
Note: This list was vetted with the assistance and suggestions of the AI, Bing, and ChatGPT-4.