Future Concept 2

I experimented more with Vizcom’s Modify and Animate features in this post. I imported my front 3/4 featured image from my last post into the Vizcom workbench. I selected Modify>New Views. On the Modify block, I picked a rear 3/4 view. I quickly had a good selection to choose from. I modified my favorite one in Photoshop to adjust the proportions. I added a reference image back in Vizcom to generate the featured image from my Photoshopped image.

I’m not focused on brand with this series. It’s all about learning how to best use AI on my ideas and speeding up the design development process. With Vizcom and Photoshop, I created the featured image in about 3 hours.

As for the design itself, I like that it looks more like a Mars rover than a car on planet Earth. I’m imagining a futuristic car by revisiting the light airy concept cars from the 60’s era. As a teenager then, I became a car nut during the halcyon days of car styling. Contemporary car design today emphasizes more mass in the bodyside with smaller DLO’s which I also admire.

I generated the AI videos in Vizcom by adding the featured images to Workbench then clicked the plus icon and selected Animate. In the prompt for the rear view, I wrote ‘car slowly drives away’. The result was amazing on the first try! I tried to generate a video in the original ChatGPT Sora with the same prompt but had bizarre results. The AI was hallucinating by producing strange twisting shapes. I decided to test Vizcom further by generating a video with the prompt ‘person gets in car with door swinging up’. I was dazzled by the results!

However, I did get a nice still image from the Sora video which I cleaned up in Topaz Labs Bloom which applied upscaling to it. I’ve included the before and after images below. Note that I still had to remove many dots or specks on the Bloom image in Photoshop.

Footnote: I was finally able to get an access code for Sora-2 by downloading the Sora app on my phone and allowing notifications. I was very disappointed with the results, however. When I tried to generate a video where the person opens the door like I did with Vizcom, I got a message stating ‘We currently do not support uploads of images containing photorealistic people’. I couldn’t get Sora-2 to have the car drive forward like the Vizcom video either. Very disappointing!!

AI is blurring the lines of what my contribution to the design is and what it creates. I become an editor of design proposals that I pick and choose from to blend in a harmonious way. Reza Bird at Vizcom beautifully summed up the current state of AI: ‘AI creation lives between intention and the control of randomness’. Maybe that sums up the human mind as well.