|
It’s funny now, but back in the '50s, this was my idea of being “on line.” A full day of lessons-learned with my closest friends and no adult supervision. It usually concluded with some incredible highlight like Billy "Leimy" Leimbach jumping from the top of a construction scaffold with only the aid of an opened umbrella. We all watched in horror as he courageously leaped. But his foot remained entangled in the metal framework where he was left hanging upside down with a broken ankle.
Now, 70 years later, “playing around online” has an entirely different meaning with the unprecedented changes in human activities since the advent of computers, cellphones, and other personal electronic devices.
Besides communicating with email and text messages, I spend considerable time watching documentaries on YouTube and generating reference photographs for future paintings using AI on Midjourney. For example, I started “The Great Northwest” painting shown above after generating 100-200 reference images of the Oregon coast with prompts to Midjourney. I always state “photographic” in my prompts because I ultimately want to add my own interpretation into a representational painting. As a maritime painter, my prompts also contain varied combinations of location, time of day, sky conditions/colors, atmosphere, tide, weather, sea conditions, view point, birds/seagulls, time period, people, type of vessels, harbors, etc. These photographic combinations yield many choices for my final watercolor composition.
I’ve never learned as much about art or enjoyed painting more than I have by using the current resources on line. And although I consider my paintings to be original representational paintings achieved after decades of developing watercolor techniques and talent. . . I realize that today's most established artists are now competing with “newbies” on AI who have never applied paint with a brush to canvas or paper.
This was recently discussed in the monthly newsletter of Peter Woolley, a professional watercolor artist and friend who lives in the United Kingdom. Peter has his own YouTube Channel with many painting episodes. He also routinely conducts painting classes on cruise ships. His next painting workshop is on the Saga cruise ship “Spirit of Adventure” leaving from Portsmouth, UK on a 30-night voyage from December 6th to January 5th, 2025 in the Caribbean.
…… ref www.peterwoolley.co.uk to register for his newsletter and receive generous painting tips
Peter Woolley’s most recent blog is below.
Artist or Text Prompt Engineer?
….. By Peter Woolley
The development of Artificial Intelligence (AI) marches on at a breathless pace and the existential threat to traditional artists becomes even greater. Or so it would seem if the news is to be believed (and let’s not even begin to start looking at the rise of fake news…).
Before I go any further, I feel I should reassure the reader that this isn’t just going to be another anti-AI rant. The fact is; AI is here to stay, whether we like it or not, and over the next few years it is going to become even more ubiquitous, in all walks of life.
As a tool, Large Language Models (LLMs) – in all their variety of forms – are, technologically speaking, nothing short of impressive, and I’m sure AI will transform many different activities and jobs as it becomes even more powerful. In the world of art, music and literature, it’s already making something of an impact. People who have never written a novel before can ask AI to write it for them, and then upload it to an online marketplace along with all the other thousands of similar tomes added each day (a staggering 7,500 per day according to Amazon). The surge in suspected AI-generated content is so great that Amazon has now introduced a cap of a maximum three uploads per day. Three novels per day? How will authors cope?
People who have never learnt to play a musical instrument have a similar set of AI tools by which they can create an entire, completely polished and packaged, musical track from a hastily-written, uploaded lyric. Supporters of technology will, of course, defend it by saying that such freedoms shouldn’t be seen so negatively. Instead, AI should be seen as democratizing; if anyone – no matter how little experience they might have (from ‘none’ upwards) – can write a novel or create a musical album from a simple idea, then that should be seen as a good thing, right?
Last year, Jason Allen, a Colorado-based ‘artist’ won first prize in a digital arts competition at the annual Colorado State Fair, with a piece called ‘Théâtre D’opéra Spatial’ – you can view the winning image here. To be fair to Mr. Allen; in his submission he made it perfectly clear that the artwork was created by Midjourney, a popular online AI art tool. He claims it took him 100 hours to write 624 text prompts in order to create and refine the image, an artistically gruelling and arduous feat I’m sure, for which he feels justly earns him the right to copyright his image. To say that this has opened up a can of worms is quite the understatement, particularly since Jason Allen’s legal battle has not exactly gone his way, with the federal copyright office turning down his appeal on the grounds that – according to them – Allen is not the ‘author’ of the piece, and that ‘his sole contribution to the Midjourney Image was inputting the text prompt that produced it‘. The Copyright Office has now rejected Allen three times over the last year, saying that Allen did not have true “creative control” of the work.
I have just a couple of things I want to say about all this.
Firstly, ‘artist’ has always been something of a divisive term, and there will never be a true consensus on its absolute definition. Jason Allen clearly considers his process of entering text prompts into an online tool to be on the same level of creativity as loading a paintbrush with paint and applying it to a canvas or sheet of paper. In fact, he would argue, I’m sure, that without his text prompts, the artwork wouldn’t exist at all (which is true). Taking this to its logical conclusion, however; when someone commissions a piece of art from an artist, and through multiple conversations, manages to encourage that artist to paint them a picture customized precisely to their personal requirements, does this not make the commissioner an artist too?
Let’s not forget that machines are fundamentally dumb and stupid (the clue is in the name of the technology – any intelligence that might possibly be attributed to it is entirely artificial – therefore not real), and that any image produced via AI software is only able to pull off that trick because it has been ‘trained’ on the vast, global library of art currently available online, produced by ‘real’ artists who have, collectively, worked for many thousands of hours to perfect their skills, and whose work has effectively been hoovered up without their permission or appropriate reference or any suitable recompence. What about their rights?
I’ve looked at Midjourney, and played with it. I’ve also tinkered with several alternatives including Kling, and have noticed that, what might at first seem quite impressive – spectacular even, if wielded properly – all starts to look a bit bland and ‘samey’. But then, maybe I’m just a little biased, and perhaps feel slightly threatened by it.
Just for the sheer fun of it, I also entered the same prompt – masquerading as lyrics – into an AI music generator (AIMusic.so), with the instruction that the result should be Acoustic – Folk – Ethereal – Female – and it spat this out…… make of it what you will.
Note: Sorry, I was not able to copy and include the audio track. (Bill Hudson)
To my mind, people who create art by feeding text prompts into an AI engine should perhaps be called something different… how about ‘Text Prompt Engineer‘. After all, there is some skill to be had from knowing exactly how to word your prompts in order to create precisely what you’re after, and those who are good at it are wanting to protect their knowledge in order to capitalize upon it. Professional Text Prompt Engineers are going to become very highly prized, and probably well-paid, in the years to come.
In my opinion (and other opinions are available); real people who have dedicated years of their lives to learn the skills required to wield pencils, pens and brushes and create great and compelling works of art – who have every reason to feel proud of their work and describe themselves as ‘artists’ – are not going to die away any time soon. The same goes for true wordsmiths and real-life, talented musicians. I predict a backlash from all this AI nonsense sometime in the future – maybe not the immediate future, it’s all too new and fresh, and exciting – perhaps 10, 20 or 30 years from now, when AI has proved just how empty and soulless it is, and folks have tried and failed to monetize it successfully (despite the fact that some will already have done just that… remember NFTs?…). True artists of all persuasions will once again flourish and prove their worth, and show to the world just how amazing and creative we can be as a species. To achieve that, however… those skills need to be passed on from generation to generation, and kids – as a priority – need to be encouraged to explore their creative sides wherever possible.
Or maybe I’m just a dinosaur and this is the future…
Sources:
CPR News
The Pueblo Chieftain
Museum World (Interview with Jason M. Allen)
Categories:Art GeneralArt Trends
Peter Woolley
| 
