Then it happened DALL-E 2and things called up to 11.
Artificial Intelligence Applications for Text-to-Image Synthesis Like DALL-E 2 Allowing anyone with imagination to create pretty much anything they want, simply by writing a few how-to words known as “prompts”. can you imagine Dramatic love story between John Oliver and Cabbage or turn around David Bowie’s words in surreal artwork Worthy of Ziggy Stardust’s own album. But, in the hands of someone as creative as Trillo, these AI tools are the equivalent of going from using a file zoetrope to own Artificial light and magic Follow all your whims using only the command line.
““It’s a very exciting and exciting time for content creators,” Trello told me via email. “On the one hand, AI democratizes the making of images so that people who speak more can express themselves visually. It also gives already visible people a way to advance their work and walk paths they may not have explored before.”
“[The experiment] It worked out much better than I expected,” Trillo says. “My next motivation was to do a clothes swap, which is also Impact I’ve practically done in previous projectsAnd the But I wanted it Try some other ideas before using it with a full-bodied person.” This is where the power of the tool became very apparent to him. The generative work with AI is so fast that “we can now access a multi-ideal universe,” he says.
Overwhelming and unlimited creative power
To create the outfits seen in the video, he used variations of the many text messages to guide the AI, from an oversized “Iridescent Purple Mylar T-Shirt” to a “Lavender Purple Puffy Jumpsuit” to a Futuristic “Purple Vintage Iridescent Jumpsuit with a Mock Turtleneck” Puffy Feather Shoulder Pads Pioneering Fashion Japanese Minimalism 2040 Barbarella.
“[It] It opened the door to some unbridled designs that I wouldn’t have come up with on my own,” says Trilleau, who describes the power as “unlimited and overwhelming.”
Another stop-motion DALL-E 2 experiment was by Trello, who also used text-to-sound AI to narrate it in the voice of Sir David Attenborough.
The difference between using traditional technology and artificial intelligence is astounding. As Trello says, you’ll literally need to design and build 100 sets, then change the model to every little frame and you’re in control of the camera. Another way is to design and make the clothes in 3D, having to create the fabrics and textures, then lighting and doing the necessary composite work on the video. But DALL-E 2 can do this through a text prompt. It not only generates something like clothes, but it can also recreate an image and instantly fitting something into that image. This is a unique feature of DALL-E 2 that other AI programs don’t: “DALL-E analyzes the original photo’s aesthetic, lighting, perspective, everything, and seamlessly blends something new into the original photo. It’s incredibly good at adding things to a scene or filling the scanned portions, a process known asinpainting,Trello says.
Create a fashion show with artificial intelligence
Actual video creation is a straightforward and even tedious process. First you have to take the base video, which is just a nice shot of his wife, Shyama, On the way to the garden. Then you have to extract the video frames and feed them one by one to DALL-E, entering text in the prompt for each one. “It’s basically AI-generated pauses,” he says.
Some fits that didn’t work pic.twitter.com/7pbIRvVZo2
– Paul Trillo (@paultrillo) August 23 2022
“Once I explored a few different directions to go, I sharpened a certain wardrobe style,” Trillo says, balancing visual design diversity with visual consistency. He ended up using 115 clothes, with countless clothes left on the cutting room floor. Those things had to be arranged in such a way that they flowed together organically, he says, but it was also unexpected. Finally used another AI program – called RunwayML – for Rotoscope Image sequence in the source video. Then, to finish the sequence, he created the floating objects that you can see with DALL-E, again using stop motion and placed them in different layers for more depth.
I’m not afraid of artificial intelligence
Trello thinks one of the dangers of DALL-E is that it’s free time. Her “imagination” is so wonderful that you can easily waste hours and hours in the process of exploring. But he thinks you can curb that compulsion, which is exactly why it’s just such a powerful tool in your creativity arsenal.
Trello knows that DALL-E could have negative effects on the creative industry, but he insists, “Tt will not take any jobs away from visual effects artists.” If anything, he expects, “it will create efficiencies for the work they are already doing. It will open the door to entirely new types of technologies as well as allow low-budget projects to have realistic visual effects.”
Which is logical. I can see how really creative people will be safe and empowered by these new tools rather than being threatened. I can imagine people who are proficient with After Effects or Photoshop, but have limited creativity, and are losing functionality, just like many others lost to technology that has enabled others to do amazing things.
Trillo makes another good point: “If everyone could make a scene, the scene would get boring.”