AI and The Arts

 

Image created by FLUX.1 Kontext by Black Forest Labs 

Introduction

As AI models grow larger and larger, companies have to find different ways to get more training data. The more diverse, rich, human content these companies can find, the better these models can get in imitating us. 

What do these companies do with these datasets? Well, they feed them to huge Multimodal Models (models that can support several different types of inputs) to make art, music, text, and much more. 

But, what does this mean for the creator? Sometimes, companies will copyright strike music from artists despite it being in the public domain because the company has a recording of it. 

In this case, can Artists and Musicians copyright strike AI for stealing their art?

How does AI Art work and how do you detect it? 

In essence, a model takes in a text input and creates a noisy image, and then tries decoding it. This is called Stable Diffusion. 

Take Flux K.1 Kontext for example (above). These images are starting to get more and more realistic. You can still guess this is AI though. For starters, the signs don’t seem to be correct and they look like gibberish. Also, the image just feels too perfect. Getting a photo in a city in the middle of the street with nobody? The snow is barely hiding any facial features? And a clear view of a huge tower? 

What’s the point of this image? Why even create it? 

Why do people make art?

Art is a creative expression– a way to show emotions, create joy, and show freedom. Under this framework, what makes AI images “art”? All they do is try to explain some idea of a human prompter, but they truly can’t capture the explanation: all they do is give weighted averages of what’s possible. Sure, AI art might be good for prototyping and seeing possibilities, but it can’t capture a very clear, one idea. Again, all it’s doing is calculating an estimation of what you might want, not an exact idea.

As one Harvard lecturer (Yosvany Terry) said, “‘the ability to react in the moment, is something that artificial intelligence can’t reproduce.’” 

So, despite this, why do people defend AI “art”?  It’s just like photography and painting: just because you can photograph something doesn’t mean painting it is unnecessary. In the same way, having AI “art” doesn’t mean it will replace real artists, rather, it could be used as a supplement, creating prototypes to understand and come to a consensus with a buyer quicker, saving both time and money. 

So, can musicians and artists copyright claim AI “art”?

Unfortunately, it’s probably not possible to copyright claim art. It’s extremely difficult to claim AI has stolen your style. Although there are some exceptions (like the “Studio Ghibli" style), the lines are super blurry and unclear. 

In fact, a recent Music AI startup, Suno called the claims forced and fake. 

But, do musicians really need to copyright strike AI? Yes, because they still feed on the musician’s success. The more diverse, rich, human content these companies can find, the better these models can get in imitating us. 

In the case of music, startups have been using unlicensed works to train their models. However, companies have decided to finally bargain with startups: allow them to license the work instead of suing them. Still, it isn’t the musicians’ choice, but it's a step towards a better future. 



 

 

Comments

Post a Comment

Popular posts from this blog

What is a Multimodal LLM?

Image
Introduction Multimodal is the ability for a model to process several different types of inputs. For example, ChatGPT can support text input as well as image input as one combined input. You could also have a model take in .pdf, .md, and input text as inputs (yes, these count as different modalities).  This allows for users to give represent different ideas in its rawest form possible, such as an image, instead of describing it. If you try to describe it, you could hit hurdles in trying to properly explain it and making sure it interpreted the description correctly.  How do Multimodal LLMs work?  Typically, these modalities are "fused" together. Here's an architecture diagram,  Image from NVIDIA In this case, we have an encoder for a specific modality (e.g. video encoder, image encoder, audio encoder). This encodes them into embeddings the LLM can understand and use. After encoding each specific modality, we then fuse them together. Then, the output is projected and ...
Read more

Top 3 Breakthroughs in Computer Vision in 2024

Image
Introduction Over the past few year, AI has become popular; however, no year matches the pace of the past few years. 2024 was an extremely crucial year for Computer Vision, with new models and breakthroughs occurring rapidly. Today, I'll be talking about the 3 biggest advancements in the Computer Vision field during 2024.  3. Vision Language Modeling Simply put, Vision Language Models, or VLMs for short, are Large Language Models (LLMs) that take an image and/or text as input. Think of ChatGPT, where you import an image and then type out a question that you have over the image. It will be able to understand both and output an answer. These are also what you would call a "multimodal model," which is a model that can understand multiple different types of sources/inputs.  Although Vision Language Models have existed for a while, they have recently become better and more accurate. However, this doesn't mean it's all sunshine an rainbows: there are still several barri...
Read more

A Mathematical Explanation of Gradient Descent

Image
Example of a close-to-optimal Gradient descent for a quadratic function. Credit: Dylan Introduction Although I wrote an article recently explaining Gradient Descent, I feel like a mathematical explanation of how it works would be beneficial not only to understand how it works, but also how you derive the functions.  Explanation First, we need a way to calculate how far off the model is. For our example, we will use the L2 Loss Function,  Let N represent the total amount of data points we have, i the iteration we are on, yᵢ represent the expected output, and ŷ represent the output predicted by the model.  Let's look more into ŷ. Since ŷ is the predicted value, this is the function that needs to have their parameters changed. Our goal is to change these parameters enough to minimize the loss of ŷ. For our example, let's set ŷ to be a linear line.  The values in "a" and "b" are our parameters. These will be the values we will have to change within f(x) to chang...
Read more