July 5, 2022

Jump takes a dive into the surreal world of Artificial Intelligence Art

Stephen Hawking said of AI:

“It will either be the best thing that’s ever happened to us, or it will be the worst thing. If we’re not careful, it may very well be the last thing.”

Well we’re certainly hoping the current craze for Artifical Intelligence Art doesn’t precursor the end of mankind!

As designers working in a creative industry though … should we feel threatened?

Your social media timelines have no doubt recently been filled with crazy, surreal images and an excited chitter chatter about AI Art.  But what is it exactly?  The Jump team and some of our friends in the industry gave us their thoughts.

What AI ART? (also called Text To Image AI)

AI Art is software … and some pretty impressive software at that. It first invites you to type in a sentence or string of words. It then goes away to do some clever calculations and returns with a range of pictures in different styles that have all visualised exactly what you’ve written.

Digital Creator, Marques Brownlee explains more in great detail here.

So typing in something simple like:

 ‘Red apple on a blue plate’ 

…and getting a picture back of a red apple on a blue plate doesn’t sound particularly mind blowing.

But what if you type something far more crazy in? (This is where the fun begins!)

‘Bearded man with teeth made from unicorn horns riding a narwhal hyper realistic cinematic epic detail.’

That escalated quicky!

Jump’s Senior Art Director, Lee Jacobs got hold of the Midjourney AI software to have a play.

“Having seen lots of beautiful AI generated images popping up in LinkedIn and Instagram, I decided to give it a go. I managed to get a login for the beta of MidJourney, running through Discord. 

I was slotted into a group of newbies that were busy feeding the bot with cool descriptions and phrases. I watched the progress of other users for a while before dipping my toe in. The images that others were creating were visually very impressive but oddly I felt a little let down by my own creations. 

MidJourney wasn’t totally nailing my brief and didn’t quite meet my expectations. My imagined fantasies weren’t playing out as I expected, whereas I was continually delighted by the art of others. Probably says more about me as a self-critical artist than it does about the AI.”

Lee’s initial experience does suggest that the software isn’t completely easy to control (and Lee is a super talented chap) – and that there is a skill to be learned in order to get the best out of it. Good news for us creatives!

Here are some of Lee’s other sentences to challenge the software.

‘Portuguese man half dugong half Mercedes down brixton high street.’

‘Epic london sports day frisbee cinematic dramatic iron city.’

 (To be fair this is pretty accurate based on previous Jump Sports Days!)

 

So why is everyone talking about AI Art a lot now?

Wajahut Shah – Technical 3D Generalist (& previously Animator at Jump):

“Since AARON the art generating AI from the late 1960s, there have been more and more artistic AIs being developed, some that require specialised open-source software like Jupyter notebooks alongside high performance graphics cards and other simple applications that can be run off a smartphone. Accessibility to AI tools has drastically improved alongside the performance of these generators which can create intricate and complex details in a fraction of the time a skilled illustrator would need.

This of course is very alluring to creative industry professionals who want to minimise expenditure and increase quality output, naturally with the gossip and generated imagery of AI such as Midjourney scouring the internet as of late, professionals want to take a crack at this tool and see for themselves how it can be adopted in production to speed up workflows, improve communication between artists and clients or even to create original ideas and imagery.

While AI is still a ways away from entirely replacing a skilled illustrator, the question is… How long can man compete with machines?”

 

What AI Art software are people using?

A quick Google search will reveal quite a wide range of options for your AI creative adventures. Here are some top picks – and how they describe themselves on their homepage. It does start to feel a little like Vision On meets Stranger Things.

DALL-E 2

DALL·E 2 is a new AI system that can create realistic images and art from a description in natural language.

MIDJOURNEY AI

An independent research lab. Exploring new mediums of thought. Expanding the imaginative powers of the human species.

DEEP DREAM GENERATOR

Human AI collaboration.

ARTBREEDER

Extend your imagination. Harness machine learning to create amazing images in seconds.

There are plenty more including Big Sleep / NightCafe / DeepAI / StarryAI / Fotor. You’re either going to be producing amazing Artifical Intelligence Art or having the best night’s sleep ever.

 

Russell Mann – Technical Director at Jump:

“I find I’m split between being totally blown away by the cleverness of such technology. What kind of computing goes into understanding how to use a colour or combine objects in reasonable ways. Online I saw an example of it being asked to draw a baby penguin in a red hat and yellow pants. So it had to understand the colours, the clothes and how to put them on a penguin. A baby one at that! But when I look more closely I’m not so sure. I guess if I didn’t have time or if I couldn’t draw myself and I needed that illustration of a penguin .. then boom! 

Maybe illustrators and artists should be more worried? Design has always required that bit more. It’s more than just the image or style, but the whole concept and its application. Of course it could be a very useful tool to aid that spark of creation and it’s quick, but there’s more to it than that for design.

Also looking at the results the penguin was not anything groundbreaking and looking at Lee’s tests, although they are cool images did it really get the brief? Maybe a beardy man with unicorn teeth and a narwhal was pushing it – but the half man, half Mercedes in Brixton is a fail isn’t it? If it creates that spark or enables you to visualise something which you personally aren’t able to, then it is a very powerful tool, but I don’t think we can start getting rid of the designers yet.”

 

Jump is a motion graphics company so it will be very interesting to see how the software develops in the future to accommodate animation.

How do illustrators feel about Artifical Intelligence Art? 

We got in touch with the brilliant artist / illustrator Matt Saunders and asked him to give us his thoughts.

(http://www.mattsaunders.ink)

“It could be a great tool to quickly springboard ideas and conversations plus it can break our normal pattern of thinking.  In the past, drawing was an essential skill for art directors ( see Harper Goff ) but from experience I don’t think that’s generally the case in the last 12 years. With a tool like this art directors can compile some rough ideas. When working with clients, illustration is always about collaboration and trying to land the ideas that are being presented to you in the most effective way. The danger with a tool like this is that art directors might get a bit carried away and illustrators become a painting by numbers service where the creative thinking is removed on both sides.

It’s also going to be really interesting to see how this affects the NFT (Non-Fungible Token) space. If you can generate millions of interesting images in a day, and consumers become numb to cool looking jpegs, how will this affect the value held by NFTs?

For me the software is an incredible tool to springboard ideas and build upon them. The software can churn out incredible images and possible ideas but it can’t string a narrative together. This is where I see the opportunity to do more with it. I would use it more like a Rorschach test to generate ideas and worlds rather than a way to create a finished product.”

Example below shows the Midjourney image offered (left) and Matt’s inspired art (right).

Image copyight Matt Saunders 2022

MS – “Automation is a threat to any industry. It can be a great help but as humans get busier and busier there is a tendency to try to save time and money by cutting corners. There are some sectors in the creative industry that are already quite low paying and these are the sectors which I predict to be the early adopters. It will be fascinating to see the results. 

I think the real danger comes when people check their imagination out and just type in what they want.  It can make the whole experience like a microwave meal. True art direction is the ability to craft a story and a vision. The danger we could be heading towards is artwork that looks pretty but has no soul and just becomes another cool thing to swipe past. I think with technology you need to see it as an opportunity rather than a threat as it’s something you can never stop. There is no point wasting energy and stressing about what you can not control. It will always keep pushing the boundaries and you just have to ride with it.” 

 

Over on Linkedin, Creative Director, Karen X Cheng talked about her experience using DALL-E 2 to create the cover for Cosmopolitan Magazine.

(Read in full here)

I think the natural reaction is to fear that AI will replace human artists. Certainly that thought crossed my mind, especially in the beginning. But the more I use DALL-E, the less I see this as a replacement for humans, and the more I see it as a tool for humans to use – an instrument to play.”

Image copyright Cosmopolitan Magazine 2022

Allie K. Miller, Global Head of Machine Learning BD, Startups and Venture Capital at AWS made a very important observation.

(Full comment on her Linkedin page)

While I’m happy to see any STEM magazine cover (the spotlight on AI and Space is fantastic!), I’m disappointed in the caption under the astronaut’s foot.

First, there is no mention of a designer. “It” was not magically made. Karen X. Cheng created the cover with the help of AI. Without Karen and the AI working together, this cover would likely not exist.

Second, it didn’t take 20 seconds to make. Karen’s personal behind-the-scenes video mentions kickoff calls and creative reviews, “hundreds” of prompts, “thousands” of options, and “hours and hours” of design.”

 

If you fancy delving really deep into this topic there are some fantastic articles on the Oxford University website.

Art for our sake: Artists cannot be replaced by machines.

Anne Ploin, Oxford Internet Institute Researcher:

“Artistic creativity is about making choices (what material to use, what to draw / paint / create, what message to carry across to an audience) and develops in the context in which an artist works. Art can be a response to a political context, to an artist’s background, to the world we inhabit. This cannot be replicated using machine learning, which is just a data-driven tool. You cannot – for now – transfer life experience into data.”

AI and the Arts: How Machine Learning is Changing Artistic Work

(By Anne Ploin, co-authored with OII researchers Professor Rebecca Eynon and Dr Isis Hjorth as well as Professor Michael A. Osborne from Oxford’s Department of Engineering.

The cover was created by designer / illustrator Alexandra Francis utilising AI methods (WOMBO Dream) being fed the title of the report – and using the results as an artistic prompt.

Image copyright Alexandra Francis / Anne Ploin.

 

The unanimous verdict seems to be that designers, illustrators and creatives are safe and that this is a fantastic tool we should embrace and utilise for additional conceptualisation.

But what are your thoughts? Have you experimented with Artifical Intelligence Art for content production in any way? We would love to hear from you. Get in touch.

 

To finish, Russell Mann asks a really great question:

“Has anybody asked it to draw the meaning of life?”