AI artwork: what’s the scoop?
by Megan Landry
Over the past few years, stunning AI-generated artwork has exploded across the internet. From 2D illustrations and 3D rendered graphics, to hyperrealistic photos that are hard to tell apart from real photography. This pivotal technological breakthrough has changed the way we create and consume content forever, and it’s just the beginning.
Pioneer organizations like Midjourney, Stable Diffusion and DALL-E have paved the way for the general public to participate in making their own unique text-to-image creations at a low cost. These image generators rely on machine learning algorithms that are trained on massive amounts of images and visual data pulled across the internet. They are experts at translating a thought into an image.
In short, users can bring their wildest imagination to life by typing in any given text prompt of what they wish to see, and with just one click, the visual result will be generated within seconds.
The first time I experimented with Midjourney, it blew my mind. I spent dozens of hours huddled over my laptop exploring prompt patterns, testing art styles, and pushing the algorithm to its limits. I was fascinated that results were never once duplicated, no matter how many times I’d re-enter the same prompt over and over again.
Images can be “upscaled” into near-duplicate copies with all the same characteristics as the originals. I’ve created many rotoscope animations with Midjourney by upscaling the same images multiple times and then playing them back in rapid succession, rendering a jittery stop-motion effect. To achieve this result, I’d stitch the images together manually, although there are more advanced databases like DALL-E that can generate video automatically. The quality and fluidity can be hit or miss, but some AI-generated “footage” is remarkably realistic.
Beyond the visual arts world, AI has also taken over text content and copywriting spaces with the help of Large Language Models (LLMs). AI chatbots like ChatGPT rely on these tools to read, summarize, translate, and predict text to generate authentic well-written content in seconds and mimic human speech patterns in conversations.
With the prowess of any given topic or theme, ChatGPT has the autonomy to write its own original poetry, songs, stories, scripts, blogs, essays, proposals, articles, cover letters, social media captions and hashtags, and anything in between.
As a lifelong creative writer who grew up wanting to be an author, I lean on ChatGPT to inspire and extend my thoughts, perspectives, and vocabulary, rather than copy & paste from it verbatim. I also ask ChatGPT for advice, tips, and recommendations, whether it’s work-related or for my personal life.
So, what now?
Don’t blink — AI tech is moving very fast! But amongst the excitement, there are some growing concerns.
Ideologically speaking, it’s all a grey area because the technology is so new. Many creators, educators, and professionals alike believe these AI powerhouses infringe upon legal territories, such as consent, ethics, patents, copyright laws, and privacy rights. Others are worried that AI algorithms could replace them and take over their jobs.
The caution is valid, but I think the relationship between humans and AI will likely be a collaborative one. It will not replace us, but it will serve us as a powerful tool to augment routine tasks accurately and efficiently so that humans have more time to prioritize other tasks. AI thrives at repetitive processes like data analysis and pattern recognition, but it lacks humanity when it comes to emotion, empathy, conscience, creativity, and complex problem-solving.
Personally, I’m a huge fan of AI so far. I feel it adds value to my career as an artist by complimenting my qualities rather than taking away from them. AI still makes mistakes, but its capabilities are continuously evolving every day. Hundreds of AI-based organizations are popping up everywhere, spanning many fields and professions from small startups to huge industry giants.