How will AI impact our work as creatives?

In a series of Linkedin posts, I asked “How will AI impact our work as creatives?” Here’s my summary from both private chats and folks responding in the posts. BTW, this is not a scientific study whatsoever, just chat, opinions and food for thought. The questions from Linkedin are at the end of the blog.

AI is good for:

• Conversation starters amongst designers and work partners. The images can open dialog to align, debate, discuss features and form.

• Augment mood and inspiration boards.

• Digital Thumb-nailing

  • Create starter “thumbnails” that likely will be modified out of necessity or desire to fit the need of the project.

• Discover pleasant surprises.

• Designing with prompts can be a great creative exercise in itself.

• Inclusive Design-ing

  • AI allows almost anyone who can’t draw to create images of things.

  • Product Managers, Engineers, Marketing, etc. people have a visualization tool.

  • Designers across disciplines can “render” in areas not their own to engage in dialog.

  • Designers who might not have the ability to draw or have physical disabilities that might prevent drawing with ease, now have a tool.

dall-e generated images

What AI does not do well, at the moment.

• Create a Vision

  • You still tell what AI to do. Your prompts are the vision, however fantastic, “a bicycle that can fly” or grounded in realism “a vessel that dispenses liquid”

• Solve practical problems

  • AI “as is” is just not there, yet. Even with the best prompts, AI is challenged with delivering physical things that capture the right feel that also directly address core problems, ergonomics, mechanical movement and transformations, etc.

• Experience Design

  • Defining, refining and inventing new product experiences. Not quite the classic set of ID skills but certainly a thing many Designers do at large.

Authorship a fuzzy line

Today we author ideas via pulling reference images from our minds, putting them on “paper”, using hand skills like sketching with pens and brushes, folding paper, cutting foam to form the idea. The effort typically results in many ideas that have to be adjusted for project goals. Your creation.

Today, via AI, we “author” prompts and get many many ideas, maybe sketch less or not at all. So in the authorship of the prompt, some would argue, we can claim the creation of the idea. Is the image completely ours or has it been co-created? As I mentioned in one of the posts, it’s much like when a design peer suggests, “make that transition go faster, think about the texture, like a pool of rippling water and try this color”. And if it works out and those suggestions move forward, they have now influenced the design and become a contributor. If the input, human or machine, makes the design better, does it really matter who the author is?  

Curation the critical moments

“the one” image from dall-e

Some say ideas are cheap, dime a dozen, even free, and the real work comes in the execution. Details make the design, says Eames. Well, there is this thing that happens between the creation and execution of the idea. We have to select the right one. Given that AI can produce hundreds of “thumbnails” in minutes, those ideas have to be curated. This is a core skill that’s just as important as creation or execution. Curation is what we do and have been doing forever and will continue to do, with or without AI (for now). Curation requires an understanding of the visual context around Brand and the Market. This typically involves some data and I would argue, requires lots of “feel”. Along with the visual curation, comes the rational options. We have to select ideas that solves the right problems, address ergonomics, have the right cost, consider Design For Manufacturing, etc. While lots of data can be gathered here, decisions are needed to move forward. Curation entails a holistic view of the product and it happens over and over from start to finish. It also includes collaboration with stakeholders to get to the best possible solution. AI isn’t not doing this…

“At the moment” AI

• does not make everyone a designer.

• is relatively inexpensive to use, if the cost becomes too high, we all might revert back to “normal”. So don’t let those basic skills atrophy.

• is not perfect and does not create usable work “as is” unless quirky and intentional error is your design intent!

• does not provide 3D CAD, what will that change when it does? :)

• interaction is not yet conversational (think Tony Stark designing his suit with JARVIS)

• And for that reason, is not great at iterating the thing it just produced.

• still takes instructions from you. That’s a good thing, yeah?!

AI will get better and it’s moving really fast. I think AI and how we choose to incorporate it in our process has been one of the biggest shifts to come our way in a while. Unlike CAD, 3D printing, Rendering SW that requires specialized training, AI is different in that just about anyone can use it. This makes the process potentially more inclusive at large. It can change the way designers across disciplines work together and alter how designers work with their business partners. AI is here to stay, regardless of what’s next, embrace it and get in front of it.

Feel free to comment below with anything I’ve missed or additional thoughts.

Questions from the Linkedin posts:

#1 Is it OK to use AI for moods boards?

#2 Is it OK to use and IA image as “underlays” or the start of a design adding “human” mods?

#3 What if a designer on your team submitted AI “as is”?

#4 What if a non-designer peer, Engineer, Product Manager, Marketing, etc. submitted AI to the concept party?