top of page

Doing Fingers Better : The Battle for Perfect AI in Art

The role of AI in art and creative fields is a topic of ongoing discussion and debate, I’m pretty sure we all have our own opinion on our AI faves or AI as our foes. While AI systems, such as those focused on generating visual content, have shown impressive capabilities, it took a million years to make fingers happen.



We Really Wanna See Those Fingers


The battle for AI in art does not solely revolve around getting fingers right. While fingers were indeed a challenging aspect for AI models like DALL-E and Clip (which powers Midjourney), the overall goal of this use case for AI is to develop artificial intelligence systems that can understand and generate creative and artistic content.

Fingers have been a difficult problem for AI models because they are intricate and have complex shapes and movements. Capturing the fine details and nuances of fingers in visual representations requires a high level of precision. Historically, generating realistic images of fingers and hand poses has been challenging for AI models, especially when it comes to representing subtle variations and movements.


AI systems are designed to learn from existing data and generate outputs based on patterns and examples they have been trained on. While they can produce impressive and novel results, they lack the ability to truly understand concepts, emotions, and abstract ideas in the same way human artists do. Concept artists and fine artists bring their unique perspectives, experiences, and creative thinking to their work, allowing them to create original and innovative pieces that reflect their individuality.


However, it is important to note that the progress of AI models in the field of Art AI goes beyond just fingers. Researchers and engineers have been working on various aspects of visual understanding, generation, and creativity. AI models like DALL-E and Midjourney have achieved significant advancements in generating complex images, understanding context, and producing creative outputs.


What the Eye Sees, the Heart Feels, Right?


LinkedIn. The cultural melting pot of flexes, opinions and the harshest critics of everything. A CEO friend from a brilliant web3 gaming company went through a phase of calling out LinkedIn AI art aficionados with the famous words “show me them fingers” each time someone shared an image of “red haired woman standing in Parisian street wearing Hermes” [replace this prompt with similar ad infinitum] but beneath the surface of these, er, surface images is perception.


The focus on fingers in certain AI projects might be due to their significance in human visual perception and expression. Fingers play a crucial role in many artistic forms, including gestures, hand poses, and fine details in visual artworks. Go and talk to a game artist, they don’t just show up fresh from a tutorial on Youtube; oh no—though they look like they do. Game artists are artists. It's just that they apply their craft in games. Their work is studied, developed, iterated upon, it isn’t just farted out through a machine. Oh and they do fingers. Sometimes, they just do fingers. And thumbs.


AI researchers aim to improve the overall realism and expressiveness of AI-generated art, of course they do, but their machines need to learn a whole lot of fingers.


Just Enough


Art is not only about the final visual output but also the process and personal touch involved in its creation. Artists use their hands, emotions, and experiences to express themselves through their work. They bring a human element to their art that encompasses more than just visual representation. This human touch and expression are difficult for AI systems to replicate fully, I mean, this is not a lie, these things are difficult to understand as a human sometimes, let alone AI. Remember that AI is not a person, you are. Which is why flexing a ripped Albert Einstein across your socials feels more like the end of days than “hundreds” of leading thinkers heralding ChatGPT as the Axis of Evil.


Is art subjective? You’d better believe it. And it can be interpreted differently by different individuals. Human artists have the ability to convey their thoughts, emotions, and messages in ways that can resonate with others on a deep level. AI systems may struggle to capture the same level of nuanced and subjective interpretation that human artists can achieve; and because art is often a collaborative process—the human connection and interaction that take place during the creation and appreciation of art are significant aspects that AI systems cannot replicate.


Instead of replacing artists, shouldn’t AI systems serve as powerful tools that augment and enhance the creative process? Yes. That might even help the finger issue magically disappear (though I believe someone already got the memo) They can be used as sources of inspiration, assist in generating ideas, or provide new tools for artists to explore their creativity. AI-generated art can also open up new possibilities and aesthetics that were previously unexplored. The collaboration between human artists and AI systems has the potential to yield exciting and innovative results, while still maintaining the unique contributions of human creativity. At the end of the day artists are bringing us soul, they’re speaking to us from a place of understanding. If you want a generative wolf howling at Uranus, that’s between you and your therapist. I’m not feeling AI art at all, I want to feel a connection when I’m standing looking at something, I want to feel that The Artist Is Present whether I’m in a game or a museum; so give me what you got and bring those fingers too.




コメント


bottom of page