The web-toon and webcomics are now drawing open-source AI, not a pen.
Hello. I’m Luke Yang, CEO of BELIVVR, an AI webcomics platform and webcomics production company. BELIVVR was a webcomic technology startup that launched a VR webcomic called Comixv in 2017. However, the VR webcomic platform, which approached the concept of ‘drawing’ comics in VR without 3D or separate tools, did not do as well as expected.
There were many reasons, but the 360-degree image format used in VR took more work to draw compared to vertical webcomics, and the narrow market was the biggest reason for the failure. We made various technology development efforts, such as collaborating with Ajou University to research artificial intelligence coloring, but due to ComixV’s urgency to survive, technology investment was slow. Alternatively, we registered a patent for learning the artist’s drawing style using AI technology and converting it to VR webtoon format.
Fortunately, WebXR technology, which shows an immersive environment on the web, which was used in VR comics, could be utilized once again as an immersive online technology during the corona pandemic, so we were able to pivot and grow into a company named BELIVVR.
Meet the Stable Diffuse That Will Conquer the World Pictorially
Then, in August 2022, I heard about the development of Stable Diffusion. Stable Diffusion is an artificial intelligence model distributed by Stability AI under an open-source license and released on August 22, 2022. I have a background as a NAVER Webtoon PD, but before that, I was an AI developer at NAVER, so I ran Stable Diffusion for the first time and was shocked by the performance of the advanced AI. At the time of the Bucheon International Manga Festival in 2022, we were experimenting with Vtuber, and I was in the corner of the booth, installing Stable Diffusion on my laptop to print out images of vvri, the character of BELIVVR and then trying out the technology to create YouTube with the characters created in Stable Diffusion.
Stable Diffusion + AvatariPy for creating and experimenting with VTuber, October 6, 2022, BELIVVR Online Conference>
When I hadn’t thought about making a webcomic with Stable Diffusion because it was challenging to get the composition I wanted.
After that, I watched with interest as sexualized images by AI caught fire online, and based on my experience that the significant changes on the internet in the past were driven by adult content, I realized that something profound was happening.
https://twitter.com/ChromedSets/status/1575373919816921088
Then, in February 2023, when Stable Diffusion started to explode and the term “HabitatMountain” was coined in Korea, I became convinced that AI was moving beyond people’s rejection and into their favor.
Most of all, I am confidentntrolnet extension, a fixed character pose. One more, I found unlimited potential in the Github-based plugin extension of Stable Diffusion.
https://github.com/Mikubill/sd-webui-controlnet
The power of open source to drive stable Diffusion
While much has happened in the intervening years, including the Novel AI hacking, a powerful force has been at the center of this change: open source. Open source has been a fundamental force in the evolution of computers, and there’s no feature on most computers that hasn’t been influenced by open source, from the cell phones and internet services we use daily to our computers.
This open-source ethos of releasing the source of SW for free and letting people fix and improve it together started with a concept called CopyLeft, which initially emerged as the opposite of copyright. CopyLeft is “a movement started by Richard Stallman in the United States in 1984 that opposes the monopolization of information and programs by capital, arguing that all programs and information should be freely shared and allowed to improve upon each other.” A significant outcome of the movement is open source.
A GitHub link can add a stable Diffusion with many extensions and new open-source features.
In that sense, open source started as the opposite of copyright, which guarantees monopoly. Copyright is a concept that protects authors, but it also works as a right that guarantees the economic rights of inventors, inventing companies, and authors by ensuring the monopoly of intellectual property rights. Especially in the case of the software industry, inventions often require large amounts of capital and effort. Hence, it’s a natural movement, and open development is a tool to promote technological advancement because it can benefit more people.
Because software is so cheap to replicate and deliver over the internet, its propagation velocity is frightening. Once these stable diffusions were open-sourced and released to the public, many developers began to improve upon them, and once the technology was open to everyone, lexica.art popped up to share related prompts, civitai.com to share various models, and communities like Arca. Live to create and share AI images utilizing these open technologies began to pour in. And all of this has happened in less than a year.
In AI terminology, there is a term called the AI singularity. The AI singularity is a term that refers to a specific point in time when AI exceeds human capabilities. In past anime and media, when AI develops, it is depicted as living alongside humans, like the AstroBoy animation, and they are shown growing at a “human pace.” However, scholars say that the rate of development of AI will not be at a “human pace,” and the moment may be like a flash in the pan. Perhaps we are passing the “singularity” in some areas; image drawings are passing it to some extent.
This high speed is because the software exhibits cumulatively increasing performance metrics and is very cheap to deliver. We have about 15,000 die-hard fans of the Stable Diffusion WebUI and 77,000 friends who are interested in it, and one of those friends is me.
developers interested in stable diffusion web UI open source
With so many developers interested in and teaching AI, you can imagine the power of collective intelligence. When one of our developers develops or announces a cool feature, it spreads worldwide, upgrading every StableDiffusion on every computer. Even as you read this, some developer is probably working on StableDiffusion.
It’s no exaggeration to say that StableDiffusion AI has already become an integral part of the AI ecosystem in the design industry, and I don’t think any global company will be able to surpass it shortly because it’s not only about the power of software algorithms but also the power of human networks. There’s Midjourney, and there’s other generative image software, but one thing is for sure: StableDiffusion has established itself in the market as the definitive solution in the global design ecosystem.
This is no more evident than with ChatGPT, a hot topic these days, and the same concerns are shared by the Google developers behind ChatGPT.
Google engineer warns it could lose out to open-source technology in the AI race.
And this open-source-driven development of AI means that AI is a human intellectual child. We don’t know how idiot or how great doing, but it’s hard to turn back now that it’s here, and we have to live with this new era. That’s why we boldly released our AI webcomic platform and webcomic made by stable diffusion AI.
loool.io:
LOOOL — Get premium AI-generated webcomics site in VR
