Here we go again.
Going back a decade, and more, there were two topics design pundits loved to blather on about: should designers program (code), and what do we call ourselves?
There were generally three arguments:
- Designers need to work in the medium for which they are designing (e.g. digital)
- You need dynamic artifact to show how the thing is supposed to work (i.e. prototype vs photoshop)
- Employers want unicorns who design and code so they don’t have to pay two people
My opinion: No. Which aligned with that of Alan Cooper, who built Visual Basic which made it easy for designers to ‘code’, e.g. drag and drop UI elements on a screen. He said designers should know the challenges a developer faces, but it’s a waste of designer skill that should be spent on the user research, etc., you know, Design stuff.
As someone who started ‘coding’ HTML in Notepad and prototyped in CSS/JS, there is a practical reason not to code. If you are a half-assed developer you will produce designs limited by what you can do (i.e. half-assed). The more time you focus learning to code to get the prototype to do what you need it to do, the less time you are doing interviews, contextual inquiry, documenting design rationale, etc.
That said, what you should gain from the development side is a sense of what is possible — so you don’t sound stupid proposing things that can’t be built, and you can judge the relative effort involved to code different design options to get things out the door quicker. You also benefit knowing how to inspect in the browser to QA front end problems. Instead of code camp, all this is better accomplished by doing a lot of discovery and trying out new apps/sites to learn current state of the industry, with basic CSS knowledge.
But hey AI IS CHANGING EVERYTHING!
AI IS TAKING YOUR JOBS! LOOK AT ALL THE AI TECH LAYOFFS! Not mismanagement, juking the stock, or tax code changes — it’s all your fault for not being AI-ified…
Designers are now being told to ‘reinvent’ themselves — bleed into the product side and dev side, and are being threatened to incorporate GenAI/LLMs into their work or get passed by. No matter that all these execs and pundits aren’t telling you how to use it, just dumping it on your desk and hoping you will figure it out to save their ass…because everyone is flailing around now.
It goes both ways — product thinks they can skip design because vibe coding, devs are always happy to ignore designers and now they don’t have to learn front-end. All the arguments for designer coding are there — except now we can jump directly to production code without the prototype!
You know what is missing in all the ‘vibe designing’ GenAI/LLM talk? Actual Design work. Identifying the problem to be solved, user flows, testing, validation, etc. UI is really the least significant part of the Design process, and already simple enough with design systems, patterns and templates. GenAI has made it so much more efficient to launch a substandard product that meets nobody’s needs. Bonus: it looks pretty much like every other app. If you start on about how we can now do AI testing with artificial participants I will slap you in the virtual face.
I grant you that GenAI/LLM has made so much of the process easier–I would never go back to transcribing audio/video by hand, or synopsizing and munging participant files together. But it is not an end-in-itself, just a tool to do what we have always done (e.g. design thinking, double-diamond, etc).
Just because now you can describe an app (hey we’re back to command line interfaces!) instead of learning C++ doesn’t change the fact design still has to happen.