 |
| Conan is angered by the more interesting social media feed of his rivals and vows revenge |
Conan! What is valued in life?
Conan: "Wahd iz desiahd, rahwr und hawhd."
You mean desired, rare and difficult?
Conan: "Yaws! Dad's eed!"
Ever eloquent Conan makes a good point.
How does Midjourney stack up in terms of value?
1) Desired.
This question is still open. It is desired for imagery on social media. It's a lot of fun to play with. Will it be employed commercially? So far, the highest profile use has been as a gimmick cover for The Economist. I can see it being used for hundreds of thousands of indie book covers (millions of those out there already and increasing daily), or for super quick concept generation. People who want imagery for their imaginary role playing worlds, games, etc. will find it incredibly fun and useful.
The engine still struggles with complex structures and multi-element scenes. Hands, faces (unless isolated), and feet are especially challenging for it and stand in the way of adoption. Are these insurmountable obstacles? I doubt it.
Will they still be a problem 10 years, 100 years, or 500 years from now? What is the ceiling for AI? Will we have Hal 9000 and Commander Datas at some point in the future? That's also kind of doubtable, but we will likely have powerful AI nonetheless. Image generation for an AI like Data would be child's play.
Eventually, I expect a LOT of the technical issues to be solved.
 |
| Conan sprucing up his social media feed with Midjourney renderings of his enemies fleeing before him |
So... desired as a fun hobby tool with usage at the periphery of commercial arts (indie book covers) for the time being.
2) Rare
It's as common as a $10 per month subscription. You need only an internet connection and a web browser (web browsers are free).
3) Difficult
It's super easy, barely an inconvenience. The skill barrier to entry is... a basic understanding of english.
How does it stack up then, in terms of value?
Individual Midjourney images have... no value at all. None! They can be freely used by the art director who rendered them (human users are essentially performing an art direction and curation role), but also by anyone else. If you alter the image, only you can use the alteration, but the base can still be used by others.
So bit of a problem if you want to brand with Midjourney output.
I've invested lots of time in the tool over the last 3-4 months, however, and I've had a ton of fun rendering pictures for stories that would have taken me YEARS to have done otherwise. It's been a blast.
Do any of the images I've made have value? As illustrations to stories I put up on my blog, or Kindle, for example, sure, they can arguably enhance the text. Beyond that, no.
Once AI renderers overcome the technical limitations they have now, they will compete at first with stock art. It will take longer for AI renderings to impinge upon highly skilled commercial artists. Eventually, some artists will move up into creative and art direction roles (as many as there are berths to take). Speed and cost are often imperatives in commercial art, and nothing is faster or cheaper than Midjourney (well, maybe stock art), so at some point down the line I expect it to make an impact.
Those doing highly idiosyncratic work may last longer. Perhaps new laws will be made to copyright a 'style'. There's nothing currently; I don't think anyone in the legislatures saw this coming. They usually lag far behind technology.
Of course, the Luddites didn't see any laws passed to protect their livelihoods...
So yes, down the line, I expect AI to be doing a lot of grunt, toss-off commercial creative work (both visual and written). Humans will curate and art direct and tweak and provide the motivation. The higher level work, complex stuff that needs to be unique, will still need to be done by humans.
Will the commercial art field expand as our tools get more powerful? Will human participation in the process continue to be necessary? How much of an appetite does the public have for art? How much do they have to spend, in both time and money?
It's possible AI and human collaborations will spark a new creative age. There may be emergent properties here that we are entirely unaware of as yet.
Honestly, I'd like to see where the AI renderers are in ten years before I really place my bets.
My current view:
• It's a super fun creative toy to play with
• It's GREAT for amateur indie writers
• It's great for world building and amateur concept art (with the caveat that whatever is made can be used freely by others)
• Midjourney imagery has zero intrinsic value
• It's so easy as to be seductive and addictive; you can waste huge amounts of time on it
• Because it is common and easy, I don't see investing time in it as truly productive. It's a crazy fun hobby; best case (for me) is that it can be used to promote / illustrate stories, or as reference for my own digital painting efforts. Whatever limitations it has in terms of structure, I do love the colour palette(s) it uses.
• Midjourney will continue to improve.
Let's check back here in another 10 years, shall we?
For now, I'm going to keep an eye on it, dabble from time to time, but otherwise I need to get back to other work already in progress... such as the new book!
 |
| Conan contemplates the social profiles of his dead enemies and reads the lamentation of their women |
All the images in this post were rendered with Midjourney V4, and have not retouched or edited. Unlike this post.
No comments:
Post a Comment