Future car concepts - experimenting with MidJourney AI
Photoshop render of Concept 1 my work using the AI thing as a base
4 years later running the image through Vizcom AI no retouch
Photoshop render of Concept 2 another iteration
4 years later running through Vizcom AI no retouch
Photoshop render of Concept 3 and a final version
4 years later running through Vizcom AI no retouch
Mid Journey AI - this is the output from MJ as you can see it’s really whacky but you can also see forms in the mess. Very interesting stuff…
I decided to revisit this post from 4 years ago with an update on the visual quality level using the current consumer grade AI models. this time I used Vizcom AI which is a much more advanced design tool then Mid Journey was / is. All I did here was take the image that I had rendered over the original MJ outputs.
I fed the images into the AI and requested a realistic render of the futuristic racing vehicle. These are the direct outputs with no retouch. There are some tty obvious issues that would need editing but it’s obvious that the visual quality has improved by a serious amount. Whether or not we agree with the ethics, I do not , AI is definitely a force to be reckoned with and its impact on the creative industries will be immense.
Original post below.
——————————————————————————————————————————
About a week ago I got turned on to Mid Journey AI by a friend and collegue of mine Matthias Beeguer you can check him out here https://www.artstation.com/sunbyte
Over the last week or so Ive been messing with this thing alongside other stuff and I must say Im impressed with the potential of this tool for fast ideation. You definately have to go back in and rework things but I found it very interesting as a way to explore form language. Especially if in the future you can " feed " it with unique image inputs, right now its text based...
Would love to hear from others exploring this...