The approach is interesting, my main take is that this type of design has become prevalent and quite standard: it felt to me like I had seen that identity on 5 different brands. With that in mind, is it really an identity if it's the same as everyone else?
Design is hard, so doing variations on a theme, or formulaic processes to get to a point where you say "I like it, it feels right, and it looks good" is a success. If you can repeat a process like the OP lays out, that's a strong foundation to build a website on.
The identity part comes from the divergence from baseline, and having a good foundation lets you experiment and "craft" easily. Nobody is gonna get a knockout web design in 30 minutes following a tutorial, but you might spend a week or three building on what you do in that first 30 minutes to achieve something unique and stellar.
If I'm not mistaken, you created some graphical assets for your brand identity.
This headline is very clickbaity.
You already created a logo and a color palette, which is almost all work done.
On top of that, the visual assets are mediocre at most. This 'brand identity' is like 1 in a thousand. This only supports my opinion that midjourney isn't capable of real creative work.
TLDR: This is not making a brand identity in midjourney, this is making a logo and color palette and getting Midjourney to create some similar images.
I was somewhat expecting this to be more of a creating a brand identity based on the company, but this is not what they did.
They took a bunch of existing "themes", but who says those are representative of the brand?
> In our case, I have already designed the logo and color palette.
So, the core brand identity had already been decided before midjourney. Then they just used the blend feature to get a bunch more images that are a mix of the 2.
What struck me is that it lacks a design vision, and so you cannot reach identity. Brand Identity isn't easy, it requires critical thinking about what/why/who.
I can see tough AI improves the baseline for the "do it yourself" people that have no design skills. But using a humanly (well-) made pre-designed template would yield a better result currently.
I am wondering if the output will improve if the user has a good understanding of what they are and how this should translate into visual language, or that the network would just be capable of discriminating between some design trends.
Slightly OT: I still don't understand why did they choose Discord to be the interface of Midjourney. So weird. It's like the early ~2010s Battlefield games where you could only launch them through the web browser + a plugin. Even if it has to be an instant messaging platform then why Discord, why not Element/Matrix or anything else?
Personally I only Stable Diffusion since there is a standalone self hosted version with Krita. But of course you need a decent PC/Mac for that which is a huge difference.
8 comments
[ 3.7 ms ] story [ 61.7 ms ] threadDesign is hard, so doing variations on a theme, or formulaic processes to get to a point where you say "I like it, it feels right, and it looks good" is a success. If you can repeat a process like the OP lays out, that's a strong foundation to build a website on.
The identity part comes from the divergence from baseline, and having a good foundation lets you experiment and "craft" easily. Nobody is gonna get a knockout web design in 30 minutes following a tutorial, but you might spend a week or three building on what you do in that first 30 minutes to achieve something unique and stellar.
Its like how everything was flat for a bit and now my phone icons are slowly drifting back to gradients
One trend people probably have noticed is that many brands over the last decade went from a script or serif font to a san-serif.
People also hate "corporate memphis" (https://en.wikipedia.org/wiki/Corporate_Memphis) which has been a trend for a while (probably on the way out but still used heavily).
There aren't a lot of unique brands out there, and if they have some kind of success they're not likely to be unique for long.
This headline is very clickbaity.
You already created a logo and a color palette, which is almost all work done.
On top of that, the visual assets are mediocre at most. This 'brand identity' is like 1 in a thousand. This only supports my opinion that midjourney isn't capable of real creative work.
I was somewhat expecting this to be more of a creating a brand identity based on the company, but this is not what they did.
They took a bunch of existing "themes", but who says those are representative of the brand?
> In our case, I have already designed the logo and color palette.
So, the core brand identity had already been decided before midjourney. Then they just used the blend feature to get a bunch more images that are a mix of the 2.
I can see tough AI improves the baseline for the "do it yourself" people that have no design skills. But using a humanly (well-) made pre-designed template would yield a better result currently.
I am wondering if the output will improve if the user has a good understanding of what they are and how this should translate into visual language, or that the network would just be capable of discriminating between some design trends.
Personally I only Stable Diffusion since there is a standalone self hosted version with Krita. But of course you need a decent PC/Mac for that which is a huge difference.