Artificial intelligence: creation versus assistance 

AI tools, stuff such as ChatGPT, Stable Diffusion, or Midjourney, are all deep learning algorithms that take user input, or sources online and learn from it, like how ChatGPT learns from the users and evolves how it interacts with them. A great case of a simple learning algorithm is from YouTubers who make them like Code Bullet, who makes them complete games like Jump King, or Tetris. A simple one does multiple actions, in Code Bullet’s case has a lot of copies of the same thing doing multi different actions, recording which action was best, and saving that as a generation, with each generation using what they learned from the last, but Deep learning AI is more complex, learning to create all sorts of responses, learning as it goes, not just replicating the previous generation.

As highly adaptable as AI is, it will never replace the role that Artists and Animators have. Everyone’s sense of imagination and style is different, and everyone’s imagination should be assisted by AI, not replaced. Not only assist your imagination, but also assist in the production process of animated shows, live-action, and movies. AI can be used for all sorts of things, like the aforementioned use, reference, and assistance in the imagination.

One application of AI is to use the art they generate as a reference or inspiration. AI can generate all sorts of references if say, you have the right prompt. Some AI software even allows you to submit your images to use as a base or inspiration for the image that it generates. An example of this is Dream. It allows you to submit images and chose a style/aesthetic of the generated image. Some studios have even published the idea of using AI during the brainstorming processes of animation, like concept art and design. Studio Mowe said. “Embracing AI has the power to bolster communication top-to-bottom between animators, agencies, brands, studios, etc. It can break down barriers to the free sharing of ideas and constructive collaboration.” AI isn’t limited to just use for inspiration or reference, but also in the actual process of animating.

The use is called interpolation, a form of this is when they increase 12 or 24 frames per second work to 60 frames. This is seen in short videos where they interpolate anime and shows. In most cases, the application of interpolation is used incorrectly, they use the AI with no human interference. No interference leaves a lot of frames that have no consistency, distorting, and have no sense of anatomy or understanding of the scene. “Current development of artificial intelligence is still in the early stages, which cannot get rid of human self-creation. In the end, it is necessary to carry on the operation input with the person the moment the subject consciousness is vague and the works of art ‘re-reproduction’ lacks the conscious creation.” The article comes from Iopscience and states that human interference should be there when an AI is used in animation to create between frames. This is in concept similar to Cacani, a program that can do interpolation. 

All though I think AI should be used to assist, others not so much. AI in essence compiles artworks and pictures from across the internet and uses them to generate art. This is in essence stealing art because the AI art has been shown to generate art with almost the same pose character, seen in Genel’s Tweet. This is an ethical issue in the AI community, I do not support the AIs doing this. The issue can be remedied by not compiling everything on the internet but by sources that are open to having their works used for the matter. But like other parts of life, AI stealing art will never be resolved, it is like crime, you can always prevent and punish but it will always happen. 

AI in its current state is unregulated. This is problematic because tools like deep fakes and voice replication have popped up, deep fakes being a way to replicate someone’s face, and voice replication is in the name. Although these applications aren’t perfected yet, they are still used for less than lawful things like extracting money from people who think their relatives were kidnapped. AI-s will never be able to be limited due to the nature of the internet but can be punished, or demotivated by making laws. The creators of the AI can keep a record of all the people who used their AI or log into the page to replicate voices or deep fakes. Those records can be made accessible to law enforcement, making them think twice before making their scam. Another way is to also have a license to create and distribute AI. This way puts a small limit but isn’t watertight, nothing on the internet is.

AI isn’t just limited to art and animation. AI has all sorts of tools and applications to assist us, some use it to study (me too), and others use it to have fun, like Code Bullet. Just think of the utility that AI can provide for different facets of life, like highly interactable NPCs in video games. Maybe an application of AI could be useful for learning languages, letting you converse in that language with an AI, picking up on your grammatical errors, and helping you learn. Of course, there are dubious uses for them like the aforementioned scam calls. AI is a remarkable innovation, but its application must adhere to human morals. This is just the beginning of the AI journey, and the possibilities ahead are truly fascinating.