UI/UX before, after and long after AI.
There are some aspects of design that are timeless. and, well, many aspects that live and die in their time.
Human perceptions, as far as the raw biology and senses and nervous system go, haven’t changed so much in thousands if not millions of years.
It’s a duh kind of statement. Except when you kibitz with enough design, product and software folks everyone thinks everything is always changing with the current season.
I could go into a deep dive of different senses and the science of perception and how artists go about solving all sorts of perceptive puzzles but that’s for a different audience. This quick post is for all the biz teams, product teams, intern teams, start up teams that just need to get some better stuff out to users.
I link to some historical surveys first. Look at em so you don’t accidentally forget 1000s of teams have argued UI/UX millions of times over. Next I give you my basic anchors for thinking about it all. Then i just link to the best of the best resources. Finally to some concepts / books that will greatly improve your knowledge in these things.
P.S. You can’t get good at design and ui/ux without doing a lot of it. You can’t get good at evaluating pictures without making and looking at a lot. You can’t get good at understanding users unless you ship a lot of software and measure a lot. Sorry I don’t have better news for you.
Every era has a Best Practices/Design Kit/You Gotta Do It This Way
But AI is different! well, its difference is that it’s more, a lot more of the same perception and cognitive challenges
if all your are doing is quick hits of info or transacting / decisioning, then yeah, AI is going to just do stuff without much interact and any AI UI will just end up being some acknowledgement of success or failure. AI kills the UI completely for all the tiny low information tasks that clog our life up.
The bigger problem for design/UI/UX in the AI age is that you actually need MORE DESIGN not less, in general. The AI can do so much already that having humans understand anything that’s going on or to shape it or retrain it needs a lot of design thinking. It’s particularly challenging because AI is contantly creating new information that has “meaning” and “intent” in it, be images or videos or audio or text. It’s not merely presenting the info/data it’s creating it.
My Simple Anchors to UI/UX
- Identify and REMEMBER the one thing you definitely want the user to comprehend or do at any given moment on any given screen/state of their screen.
- Mobile first and AI second, no matter what industry or user segment. Everyone has a phone on them all the time and it’s by far the most used common computing interface. and it will never not be for the next couple of decades.
- No one is an enterprise user, except for people with the jobs where they don’t make actual buying decisions and have a laptop that has no access or rights anyway. Everyone else, and I mean everyone, is a mass market consumer user of software, computers and devices.
- Above the fold still matters.
- No software exists or is used alone, in isolation, without a million competing contextual factors. Know your users attention field or get killed by it. What else is the user doing when they use your software? what time of day are the using it? how tired are they? are they alone? on the go?
- Users have bad information memories but long emotional memories. If the software frustrates they will remember it forever, even if it’s useful or as the secrets to the universe on it.
- People don’t read. At best they scan, but only after they’ve tapped, clicked, scrolled, thumbed, swiped. People forage and fidget, they do not concentrate until preception gives way to perception and maybe then cognition will start to work.
- 3 dimensions of data at max on a given state of the screen at once. Think about everything as a map. Humans are great at two spatial dimensions and 1 time one. You give them 1 semantic dimension then you gotta lose a spatial or the time. And so on.
- All of the above doesn’t matter unless you a/b test against number 1, constantly all the time. Hundreds of times a day. and if you don’t have 100s of user sessions a day then you really will never find out if you have good or bad UI. Your idea isn’t even capable enough of getting to statistical relevance.
Here are examples of stellar UI and information design.
Seeing Theory is one of the most brilliant integrations of text explanations, data display and interaction ever created. Simply hard to beat this level of integration and coherence.
Selfiecity is another stellar coherence between different modalities of information.
Hard to beat D3 gallary and Flourish Data Studio and NY Times Graphics… and a few others…
Other Resources
Here’s some timeless concepts/research/books I’ve found useful in my couple o decades doin stuff:
