should we use ai to replace artists?

What are the implications of using AI for creative tasks? 

The use of AI in the art world has been a controversial topic as of late. This particular image below, which won the Colorado State Fair contest, was revealed to be created with an AI program called, MidJourney. It was not specified that it was created with AI when hung for judging, however it was stated in his submission.

Image Created by Jason Allen using MidJourney 

This raises the question of whether or not AI should be given credit where credit is due. AI can now create images, poems, music, code and even articles. For those who are not artistically or creatively talented, the use of AI can be a great way to get their ideas out. Individuals can write a few prompts and have hundreds of images, poems, or snippets of code created in various styles to help inspire us for creative processes. Here are a few images below that were created with DALL-E 2, a text to image artificial intelligence model. These were rendered in less than ten seconds with the following prompt:

“A secret alien society living beneath the crust of the moon, digital art style”

Images Created with DALL-E 2 showing text to image capabilities

What are the implications of this? What does this mean for artists? Is it the end all? It’s truly survival of the fittest. Either artists have to offer something that AI cannot do or artists have to use new technology to their benefit. Those who become experts at navigating these AI text to image models will be able to use this as a new tool in their toolkit.

However, at the same time, it feels inherently wrong to tell artists to use these tools to create art. Although it may be more efficient to use AI models to create art, it takes away from the originality of the artist. Additionally, classifying AI art into its own category takes away the artist's unique perspective. It almost feels like Newspeak from the book, 1984. It limits how artists and individuals creatively think.

There is another issue. If we did classify AI art into its own category, how would we truly know if an artwork was created with AI, unless it was specifically stated.

What did you think of this article? Were you able to tell that this article was co-written with GPT-3?

Here’s the original stream of consciousness that I wrote in 7 minutes.

What are the implications of using AI for creative tasks?

There has a been a wave of conversations around the use of AI in the art world. This particular image recently won the Colorado state fair contest: https://impakter.com/art-made-by-ai-wins-fine-arts-competition/#:~:text=An%20artwork%20made%20by%20Artificial,used%20to%20compete%20in%20competitions.

It was only revealed that it was created with AI after winning the competition. With simple text to image, anyone can create what’s in their imaginations. For those of us that aren’t artistically talented, this is great. People can write a few prompts and have hundreds of images that are created in various styles to help inspire them for creative processes or get their ideas out on paper.

However, what are the implications of this? From the art scene, this technology has the possibility to wipe out some sectors. From a cartoon artist I spoke with, he mentioned he lost three clients the day after DALL-E was released. 

This is sending ripples to the art world. How can Ai be better in some sense in creating? AI was primarily used for classification problems and now it’s created with the advent of generative adversarial models created by Ian GoodFellow in 2015.

Since then, many generative models such as images, sound, music, and even 3D model generation is coming out of the cracks. What does this mean for artists? Is it an end all?

It’s truly survival of the fittest. Either artists have to offer something that AI cannot do or artists have to use new technology to their benefit. Those who become experts at navigating these AI text to image models, will be able to use this as a new tool their arsenal.

However, at the same time, telling artists to just use these models to create art feels very “group think” mentality. It loses a sense of originality. Sure you needed to original ideas to create a prompts, but it feels almost like AI art should be classified in its own category.

But how would we truly know if a piece of artwork or an article was original or created with AI? Would you be able to tell if an article was written by AI? What did you think of this article? It was co-written with GPT-3.

This article was reworded and re-written from my original stream of consciousness draft I wrote in 7 minutes.