Erich Archer had unveiled his latest project at the recent Demo Day - a documentary that tells the story of his great-grandfather, and the entire movie was written and generated entirely by AI. [You can watch the recording here]
And yesterday, OpenAI released a program called [Sora]'s new visual effects model is a big hit.
But before we get to the latest technological advances, I'd like to introduce you to Rufus Archer, who lived between 1812 and 1909. Erich, with the help of a few tools, has "brought back to life" one of his contemporaries:
- ChatGPT to write the script
- Perplexity conducts a study of events during the time period of Rufus' life
- Midjourney creates visuals, including bringing two old photos to life!
- RunwayML Completes Conversion from Still Images to Motion Video
I have a few thoughts:
- To produce such a project in the traditional way, Erich might have to pay to hire a team and spend hundreds of hours.
- But the project only took him 20 hours and about $200.
- The completion of this project has been placed on local television stations and even the PBS television network for broadcast.
- It compares favorably with the work of award-winning documentary filmmaker Ken Burns.
- The development of AI will only go further and faster.
That's just what happened on Monday.
Next, Erich will apply Sora, a revolutionary video generation model released yesterday by OpenAI. i've given an example below of a video created entirely from textual cues:
Some of the video generation models currently on the market show pretty good performance. sora is way ahead of them!
Sora is not yet available to the public, but once it is released, guess who will be immediately ready to build prototypes with it?
When you construct your innovation cycle, you're not just constructing an idea, you're forming a complete pipeline full of ideas.
Each idea evolves into a prototype, giving you the opportunity to learn new skills and the chance to network with others in the AI field and in your industry.
This means that your AI skill level will increase and you will be exposed to new opportunities as you become known in your field.
Once the innovation cycle is initiated, it is self-reinforcing. Each idea makes the next one better. And when you get a huge advance like Sora, you can just use it because you've already figured out the "basics" through your previous experiments.
In fact, the AI image recognition tool released in March 2023 [EveryAltA similar thing happened with ](). About 8 months later, OpenAI released GPT-4 Vision, which solved almost all of the problems we encountered in earlier image recognition models.
If I had just taken EveryAlt's idea and sat on it without building an early prototype, I might have been disappointed with the release of the new model-- I'd feel like my chance had slipped away.The
Again, if Erich hadn't built his first and second generation prototypes in December and February, he might have suffered a Sora hit.
But because we invested and started building, we were able to capitalize on the opportunities of new AI advances instead of being overwhelmed by them.
That's why I'm so excited about the upcoming year -- for Erich and the more than 800 other IWAI students who are working on prototypes and forming a pipeline of ideas.
In two months, you'll have a chance to see what we're working on at our next Demo Day, scheduled for April 2nd. Please mark this date on your calendar and I'll be back in touch with more details closer to the date.
In the meantime, don't miss [Sora's demo]
