Using AI (artificial intelligence) as an Architectural sketching assistant…

Or maybe a more appropriate title should be: the good, the bad and the ugly.

The good: AI has the extraordinary ability to extend your capabilities; it’s super fast; options and alternates can be output in mere seconds, and a lot of the imagery is at least compelling to various degrees as a standalone image. There is actually a lot of value to having a tool like this in your ‘grab-bag’ of ways of extending and enhancing an initial sketching/concept workflow. Architects and Designers have a great grasp on how a simple sketch works spatially, volumetrically, in 3d, etc., while most people really struggle to visualize what the heck the ragged squiggle of lines mean. AI has the ability to quickly add the missing dimensions that an architectural line sketch lacks, like a sky, ground, grass, background mountains, flowers, rocks, etc. that makes a concept sketch come alive and start to become real and exciting to everyone, you don’t have to be a designer to understand.

The Bad: See above ‘The good’. Ha ha, only partially joking. While AI is very compelling as an image generator, you relinquish an incredible amount of control and direction over the produced/output imagery. Sure, you can get very specific with the Prompts you use to get closer and closer to something in the neighborhood of the idea you were thinking of, but this begs the question, how much time is it really saving to get a worthwhile output when you could get a closer approximation with traditional tools/media? the answer? A lot of time, AI is ridiculously fast! And really good! It pains me to admit not only how much time that can potentially be saved, but how quick and how well you can iterate design ideas, and for me, that’s also the bad, as the element of struggling with design ideas and working through them was always part of the creative process….but now…a lot of people will have to seriously rethink entire notions of workflow and creativity.

The Ugly: The age old saying of ‘garbage in = garbage out’ is 100% accurate here, in all honesty, it’s more about training humans to understand how to get along with AI then the other way around. AI is what it is, humans just have to find the best way of having it benefit their own workflows, process, etc….more to come on this subject in the future, as I’ll keep pushing how to implement AI into my own creative process, so stay tuned!

Example below: The 10 second sketch test. I did a quick 10 second sketch idea of a house cantilevered on a hill, using an AI sketch to AI tool, wrote very simple text prompt for AI and 30 seconds later, had a compelling image that while super rough, conveyed and relayed an emotive, visually compelling image that frankly, a rough concept sketch just can’t provide for the average person. It’s a new world, I’m trying to keep up.