Casebook

I tried the AI Art Generation tool and it is amazing.

It was late at night on a Saturday. I was on my night shift and came across the midjourney website and tried the tool. Here are my thoughts

Melvin Raj

Sep 30, 2022 — 6 min read

I Tried AI Art Generation Tool | Melvin Raj

Without a doubt these tools are amazing. I mean, who doesn't want to type out what they imagine, and the picture is presented within minutes? Interesting right? Okay, if you get curious about fascinating ideas and how this happens, this article is an interesting one.

First, let me address the main point — AI art is nothing but you type a prompt or an idea of what your image must look like and it will generate a few variations of images based on the prompts.

You can later make changes and bring out different variations or enhance them. These AI art generation platforms were recently launched (less than 6 months ago I think) and it has at least to me a curious technology.

I first heard about this AI art generation tool on Twitter. It was a platform called Dall-E (which is now open to everyone, I believe) and I was intrigued. But this platform was providing access on a limited basis. When we request access, it took around 2 weeks for others to get access, but I didn't get and it has been months at this time.

After months, I got to know about this other tool that does the same job and took the opportunity. It is called Midjourney. It is similar to Dall-E however, you get only 25 prompts to try, and then you have to pay for creating AI art using subscription models.

Okay, For now, enough details let's get into what I put as prompts and what images were generated. ready?

I don't know why but I felt fight club with a poetic and philosophical twist should be my first prompt. We all make mistakes, let's move on.

The images that were generated are:

**fight club with a Poetic and philosophical twist | Melvin Raj**

I will say not bad because the face images look like Brad Pitt and Edward Norton's faces mashed to make a poster, which is good because if you see the movie (spoiler alert) they are two characters of one person. Alright, moving on from one disturbing visual to another.

In this prompt, I typed in landscapes with beaches, mountains, and also with a futuristic city in the far distance

landscapes with beach, mountains and also with futuristic cities in far off | Melvin Raj — **landscapes with beaches, mountains, and also a futuristic city in the far distance | Melvin Raj**

The result is kinda okay but pretty close to what I typed in. If I would have upgraded or clicked on more variations, I could have gotten an image I am most satisfied with. But, I left because I honestly didn't like it. Moving on...

The next prompt was something interesting: Indian monks meditating under trees, near waterfalls, on top of a big rock

**Indian monks meditate under trees, near waterfalls, and on top of a big rock | Melvin Raj**

The images were too satisfying to watch. It looks like someone painted these.

After watching these, I was convinced that there will be stop-motion movies from the ai generated art in the future. Imagine novelists having an opportunity to narrate a scene and they get such images and we get a graphic novel version (OOOOHHHHHH Forgot to type in something harry potter themed🤐🤐).

Lastly, I was scrolling Instagram and found these reels about ecosystems in a bottle and I thought what if we lived in an ecosystem with plants and greenery inside bottles? So, the last prompt was: People living in a huge round ecosystem with greenery

**People living in a huge round ecosystem with greenery | Melvin Raj**

It is really fascinating to see the words come to life in a picture. But, there are a few questions I had in mind.

As people are debating online that in the future, these ai generation tools will be replacing artists and their jobs.

Well, I don't agree with this. I mean yes, people may opt for AI-generated art as it might become cheaper but art is subjective and it solely depends on the person buying. It may take a while (probably years) for AI to generate art precisely as the prompt type, but even if it did, it will miss the feeling of being authentic for some reason.

Do you know what it could replace? The websites like Pexels, and Unsplash; offer free images from photographers and it is a real possibility we get options to generate images for our blog posts, youtube thumbnails, and b-roll videos.
AI starting to make realistic images like this (#Dall-e). how do we tell the difference online? I mean reverse image searches are present but we all know false news spread faster than anything. On average, false news stories are 70 percent more likely to be retweeted on Twitter than true stories. It takes true stories about six times as long to reach 1,500 people as it does for false stories to reach the same number of people (Source). I am curious; will there be new service tools for verifying if the image in the news is fake or real? Only time can tell.

These concerns can and will be addressed in the future or I may be just thinking too much on this topic but these were the questions I had.

My concluding points are, generating art from AI is fun but people do love the feeling attached to making art.

People make art to remove the stress of their busy work day, they make art to feel inspired, they make art to listen to their thoughts, and I recently came to know of this but people consider art therapy to improve their self-esteem, self-awareness, and also to improve cognitive and sensory-motor functions.

For further reading:

So, imagining AI replacing artists is just like me in 2009 thinking the 2012 movie was going really happen in 2012 (I was shit scared as I was around 12-13😅😅😅😶, and I know you had this thought too).

If you want to explore the AI art generation tools, please check out these two (I am aware of these two so adding them here)