What is AI?
AI stands for Artificial Intelligence and is a computer program that is designed to mimic human intelligence. AI is not just about computers, but rather a way of thinking. AIs have been around since the beginning of time, but they were always considered machines. Nowadays, we consider them as a type of software.
How does AI work?
AIs use artificial neural networks (ANN) to learn. ANNs are a group of algorithms that mimic how our brains work. Our brain is composed of billions of neurons that communicate with each other via electrical signals. Our neurons receive information from their neighbours and then pass it along to others. In a similar manner, AIs also do this. When they receive information, they compare it to what they already know and store it away if it matches. If it doesn't match, they add it to their knowledge base. Over time, this creates a network of knowledge that can be applied to different situations.
Why should I care about AI?
You might think that AIs are only useful for robots or games. However, AIs are actually being developed for many purposes including healthcare, education, and even agriculture. One of the biggest uses of AI today is in farming. Farmers can now use AI to help them make decisions about their crops. AI was also used to help create the Moderna vaccine during the COVID-19 pandemic. After multiple simulations in just two days, the AI provided the best solution for the pharmaceutical company to go forwards.
Are there any downsides to AI?
Of course! Just like anything else, there are pros and cons to using AI. Some of the downsides include data privacy and security issues. Also, AIs are not perfect. Sometimes they get things wrong.
AI-based Art
Artificial intelligence (AI) is a branch of computer science that focuses on building intelligent machines that work and react as humans do. In the field of artificial intelligence, machine learning is a subfield of AI where computers program themselves based on experience. Machine learning is a type of AI where a computer system learns without being explicitly programmed.
Deep Learning
Deep learning is a subset of machine learning that involves training neural networks, especially deep neural networks. A neural network is a mathematical model that mimics how neurons in the brain interact. Artificial neural networks are computational models that mimic certain features of the human brain. Neural networks have been successfully applied to speech recognition, handwriting analysis, object recognition, translation, bioinformatics, natural language processing, robotics, autonomous vehicles, game playing, stock market prediction, and many other tasks.
DALL.E 2: OpenAI
Have you ever seen these art pieces before? Do you recognise the style? Would you be surprised if I told you that I created them through artificial intelligence, in about 30 seconds…
A powerful new form of artificial intelligence has burst onto the scene and taken the internet by storm over the past few months: text-to-image AI.
Programmes such as DALL.E 2 and Midjourney are at the forefront of this budding technology, and are developing at an incredibly, and somewhat scary, fast pace. They’re built by pulling millions of images from the web, teaching algorithms to recognise patterns and relationships in the images and to then generate new ones in the same style. This means that the more images that are uploaded online, including these AI creations, the more the programmes learn and get better.
Watch DALLE.E 2 Explained: https://www.youtube.com/watch?v=qTgPSKKjfVg
Here are a couple of examples from my own DALL.E 2 account to show the variety of compositions it can produce…
Using the prompt “an impressionist oil painting of deer running through a forest”
Using the prompt “sliced orange in a blue studio set”
Using the prompt “a sunlit indoor lounge area with a pool with clear water, next to a big window, with a landscape view, digital art”
How is DALL.E 2 being used commercially?
Though this trend has predominantly been used as a hobby by amateur creatives, we’re slowly seeing it move into the commercial world, with companies using these programmes to promote their products.
In July, Heinz released an advert showing that when they put the word ‘ketchup’ into an AI art generator, it always depicted a bottle in the shape of a Heinz product. On their YouTube channel they said, “with new artificial intelligence text-to-image programs taking over the internet, we wanted to find out what A.I. thinks “ketchup” looks like. So we used A.I. to generate images of ketchup on DALL-E 2. The result? Just like humans, A.I. prefers Heinz.”
(https://www.creativebloq.com/news/heinz-ai-draw-ketchup)
Though many people have debunked this, stating the programme doesn’t actually produce bottles of Heinz, it goes to show how popular this technology is becoming and how it can be used to target younger generations. “With A.I. imagery dominating news and social feeds, we saw a natural opportunity to extend our 'Draw Ketchup' campaign—rooted in the insight that Heinz is synonymous with the word ketchup—to test this theory in the A.I. space," says Jacqueline Chao, senior brand and communications manager at Heinz.
Watch the video here: https://www.youtube.com/watch?v=LFmpVy6eGXs
DALL.E 2 was even used to create the first-ever magazine cover generated by AI. The cover was used for Cosmopolitan, an American fashion and entertainment magazine for women. The project was made collaboratively by Cosmopolitan editors, OpenAI's workers, and Digital Artist Karen X. Cheng. After hours and hours of trying to get the perfect image, they ended up using the prompt "wide-angle shot from below of a female astronaut with an athletic feminine body walking with swagger toward the camera on Mars in an infinite universe, synth-wave digital art”.
(https://www.cosmopolitan.com/lifestyle/a40314356/dall-e-2-artificial-intelligence-cover/)
Though many people fear creatives may be replaced by AI technology in the future, Cheng notes “it’s not just typing in a few words and BAM magically you have the perfect image. For something like this, there was a TON of human involvement and decision making.”
Posting the cover to her Instagram page, Cheng also notes that “women are underrepresented in the field of AI. Cosmo is an opportunity to get AI in front of women who never would have known about it otherwise. This is a field that will be responsible for so much of the infrastructure on which the future is built, so we need to make sure women are part of it.
Copyright controversy
When this conversation turns to debate surrounding what it means for the future of creatives, most assume the discussion surrounds whether it will replace artists. But a growing issue surrounds what it means for current artists and their work. The open-source programs are built by collecting images from the internet, often without permission and proper attribution to artists. As a result, they are raising tricky questions about ethics and copyright. Concerns over this have caused many photography agencies to remove images created by artificial intelligence models from their databases.
Greg Rutkowski, a Polish artist well-known for his fantasy style art, features as a prompt thousands of times in the Discord of the text-to-image generator, Midjourney. Rutkowski says he doesn’t blame people who use his name as a prompt. For them, “it’s a cool experiment,” he says. “But for me and many other artists, it’s starting to look like a threat to our careers.”
Currently, artists don’t have the choice to opt into the database or have their work removed from it.
So, what does this mean for the future of creatives?
There’s no doubt that AI generators will inevitably evolve into something much bigger in the future, but whether it’s for the better or for the worse remains to be seen.
Personally, I don’t see this as a threat to the creative industry, but as a new and exciting tech tool we can use to explore new creative territories.
Midjourney: Art Created by Artificial Intelligence
The main difference between DALL.E 2 to Midjourney, is that the latter is connected to the server ‘Discord’, where a community of people already collaborate on their creations. This gives Midjourney a bigger social aspect for users to share ideas with each other.
To start using the ‘Midjourney Beta’ you will first need to set up a Discord account, this is where you will be able to give text prompts to the ‘Midjourney bot’.
Once you have signed up for a free account (other options are available) on the Discord website, you can use the application in a web browser or download the app for Windows, Mac, Linux, Android, and iPhone.
Once you’re all set up with Discord, you can then visit the Midjourney website and select “Join the Beta.” this will take you to Discord, where you can select “Accept Invite.”
(SOURCE: The Economist featured an image created with Midjourney on its June 2022 issue cover: Twitter Link)
We have listed some research links below for a full in-depth guide on how to use ‘Midjourney Beta’ so you can begin your own creative AI-assisted journey:
- The ULTIMATE Beginner's Guide to MidJourney AI Art
- Using the same prompts in Midjourney & Dalle 2. Which one's better?
- End of An Artist's Career? A.I. Creates Art for You! - Complete Guide on YouTube
Website: https://www.midjourney.com/home
How is Midjourney being used commercially?
Listed below are some of my favourite artists looking at using Midjourney for commercial marketing solutions, although most of these designs shown are unofficial creations.
Nike x Midjourney By Alejo Bergmann
Mid Journey Ai Art Collection By Jeff Han
Via Richard Rosenman Creative Director
The fast-fashion industry is becoming a serious environmental issue, particularly as brands continue to push their products out via micro-influencers (young teens who don’t get paid for brand promotion as they’re building their own audience which they can monetise at a later date). With this in mind more and more legitimate brands are trying to find new and inventive ways of standing out from the crowd on social media platforms.
Having the availability to deliver an eye-catching, yet familiar image, that draws the attention more than most due to something not being quite right is a bit of a game changer. Especially when it comes to capturing an audience who's more than likely scrolling very quickly on a mobile device.
How is AI Working with Architecture?
Design generated in Midjourney AI by @nai9.art - “Relax and unwind from the cares of the world in this relaxing space that imitates waves sounds to relax you”.
Mood Boards for Game Design
Arnaud Atchimon on Linkedin - “I've been using it on Dreamstudio to build the atmosphere of my fictional video game Dynamo, and the changes are genuinely perceptible.
Prompts used: {heavy rain} {night} sustainable city roads || african city cyberpunk 2 v motorcycle} by marc newsom , {sunrays}, canon eos c 300, ƒ 1.8, 35 mm, 8k, medium - format print,
Then added words like: {fog, heavy fog, sunrays, rain, drizzle, heavy rain} - {electric} - {car, race car, motorcycle, bicycle}
As a part of the Midjourney beta, users who are currently participating will receive approximately 25 free images. After that, you will be required to pay either $10 per month for 200 images per month or $30 per month for a standard membership, which grants unlimited access to the service.
- Midjourney User Manual / Quick Start
How Midjourney is Being Used Alongside Architectures 3D Design Tools
“Arturo Tedeschi is an Italian architect and computational designer internationally renowned for his sculptural and visionary approach combined with research on advanced design methods, materials and fabrication technologies. His use of digital technologies blurs the line between disciplines and emphasises the semantic and emotional values of objects which engage all the senses”.
Architect & Computational Designer - Arturo Tedeschi | Instagram | Linkedin
When professional designers are using the tools of their trade, the only limitations to their creativity are their own imaginations, plus the laws of physics. With programmes like Midjourney, architects should be able to push the boundaries of the landscape even further, with no limitations on the initial concepts, then after introducing the software of choice for ironing out any problems with any physical aspects of the buildings, we should be left with some highly imaginative buildings and landscapes designs going forwards.
AI Portrait Images
Brain Castleforte Freelance Motionographer - Founder / Creative Director at Castleforte Group
For my very first attempt, I focused on one of my clients in the windscreen repair and replacement services sector.
Plus, another creation below I made in my spare time, the final character is the upscaled version chosen from image one of the four above.
Prompts: female android wearing body armour "white ceramic shell", holding a sword, unreal engine, 4k. The final product came out great considering I've used minimal prompts and only included one prompt taken from elsewhere (white ceramic shell). As well as this, it was also my final free image credit just as I was about to hit the upscale button!
Time to either take up a paid subscription? Going forwards I will make most images full length, include background prompts to set the scenery, and then continue to build the picture from there. If needed I could even import the image into Photoshop for final touches.
Another sector I will be looking at in my personal time outside of robotics and androids is architectural design and possibly transport design.
Conclusion
I have seen a lot of Midjourney images for portraits like the ones created by Brain Castleforte shown previously in the images above. I presumed this would crossover into the NFT space just as thousands of people were looking for digital artists to create thousands of images for them to monetise. However, with that bubble seemingly bursting, it will still be interesting to see how this will all play out with the introduction of the Metaverse and Web 3.0 just around the corner from the mass market to take any financial opportunities they can by using decentralised applications.
Sources:
Midjourney Website: https://www.midjourney.com/home
- The Economist (AI magazine cover)
- Nike x Midjourney By Alejo Bergmann
- Mid Journey Ai Art Collection By Jeff Han
- Richard Rosenman Creative Director
- Midjourney AI by @nai9.art
- Arnaud Atchimon on Linkedin
- Architect & Computational Designer - Arturo Tedeschi
- Freelance Motionographer on Linkedin Brian Castleforte
Useful Links:
- Cleo Abram - The REAL fight over AI art (YouTube)
- How to Geek - How to Create Synthetic AI Art With Midjourney
- CNN - AI Won An Art Contest, & Artists Are Furious
- analog_dreams - The Ultimate Beginner's Guide to Midjourney AI Art
Dimitris Katsafouros - Using The Same Prompts in Midjourney & Dalle 2. Which one's better?
