For a long time, there was an uncomfortable economics problem with UX. Many organisations understood that usability mattered, but doing UX properly meant bringing in specialists and the resulting cost. For a small company or a product with a limited budget, the choice was often straightforward: pay for UX or do without it.
That choice is changing.
AI can now review an application and provide surprisingly useful usability suggestions. It can identify obvious problems, suggest improvements and help someone who isn't a UX professional avoid mistakes they might otherwise have made.
This doesn't mean AI has suddenly turned everyone into a UX expert. It means something more practical: the cost of getting to a reasonable level of usability has fallen.
The WordPress theme effect
We've seen this happen before.
Consider WordPress themes. If someone wanted to create a corporate website from scratch without a designer, there was a good chance they would make basic visual and usability mistakes. Typography might be poorly chosen. Text might be too small. Contrast might be inadequate. Spacing and hierarchy might not work particularly well.
A good theme solved many of those problems before the person building the website even thought about them.
While you didn't suddenly become a designer, you started with a set of decisions that had already incorporated a considerable amount of design knowledge.
Good defaults don't replace expertise. They make expertise less necessary for problems that don't require much expertise.
AI is doing something similar for usability, but in a more active way. Instead of simply providing a predefined design, it can look at what has been created and suggest changes.
The result can be surprisingly good for a normal product following a relatively normal flow.
Vibe coding is another example
You can see the same effect in the growing number of applications being created through AI-assisted or “vibe coded” development.
People are building things today that they might never have built before because the cost and effort of getting from an idea to a working application have fallen dramatically.
Not all of these applications are good. Some have obvious usability problems. But that's not really the point.
The important change is that a person who previously had no practical way of creating an application can now get a functional product, along with a reasonable level of interface and interaction quality, without assembling a large specialist team.
For a small organisation, that matters.
The relevant comparison isn't necessarily between AI and an experienced UX professional. Sometimes the comparison is between AI-assisted usability and having no UX input at all.
For many small-budget projects, getting from poor usability to reasonably good usability may be more valuable than waiting until there is enough money to pay for a comprehensive UX engagement.
But “good enough” depends on the game you're playing
This is where the argument needs some qualification.
A usable product isn't necessarily a great product. And a great product isn't necessarily the right product.
There are niche products and complex applications where usability requires considerably more thought. Enterprise applications can involve complicated workflows, unusual business rules, competing stakeholder requirements and edge cases that don't appear in a typical application.
A WordPress theme that works beautifully for a simple corporate website isn't going to solve the problems of an enterprise application with a complex workflow. It can provide a starting point, but significant customisation is required.
AI has a similar limitation. It can take you surprisingly far when the problem is familiar and the flow is conventional. The further you move away from the conventional, the more human judgment matters.
And there is another factor: competition.
“Good enough” is relative.
If you're building an internal tool that only a handful of people use, a competent baseline may be perfectly adequate. If you're competing for millions of users and your competitors provide excellent experiences, good enough may not be good enough for very long.
The market doesn't value design or UX uniformly. It values the difference a particular level of expertise makes to a particular problem.
That is why there will always be a place for higher levels of UX expertise. The question is not whether expertise has value. The question is whether the problem you're solving justifies the additional investment.
AI — The Ace Up My Sleeve
With AI taking over a large chunk of the basic usability and UX work, you might think the advantage is shifting away from experienced UX practitioners. But the deck isn't stacked against me. I have my own ace up my sleeve: I'm using AI too.
When I have to create something, I don't start with a blank page anymore. I ask AI to create a first version and use that as the base on which I build.
If I need to get to ten, I don't have to start at one. I can start at five.
That may sound like a small difference, but it changes how I work. AI takes care of some of the initial thinking—the obvious structures, alternatives and basic decisions that would otherwise consume time and attention. I can then spend that saved capacity on the things that actually require experience and judgment.
Is the workflow right? Are we solving the right problem? What happens when the user does something unexpected? Where can we make the experience significantly better rather than simply acceptable?
That is where I find AI most useful. It doesn't have to produce the finished design for me. It gives me a head start.
If I need to get to ten, I don't have to start at one. I can start at five.
So while AI may be swallowing up some of the work that UX practitioners used to do, it is also swallowing up some of the work that gets in the way of doing our best thinking.
And that is a pretty useful ace to have.
The UX starting line has moved
This is why I don't find the question “Is AI taking away the work of designers?” particularly useful.
Some work will certainly disappear or become automated. As AI improves, the amount of work that requires human intervention will shrink in some areas. There is no reason to assume that today's boundary will remain fixed.
But there is another effect happening at the same time.
People who previously couldn't afford to get design or usability help can now get some of it. People who already use professional UX can use AI to get further, faster.
That is not simply substitution. It is also an expansion of access.
For organisations that have historically looked at UX as an expensive luxury, this is probably the most useful way to think about it.
You don't necessarily need to commission a large UX programme to make your product better. You can start with AI. You can use established frameworks and good defaults. You can get a baseline level of usability at a cost that would have been difficult to imagine a few years ago.
And if your product is complex, your users have unusual needs, or you're competing in a market where experience is a differentiator, you can bring in deeper expertise where it creates the most value.
The opportunity isn't to choose between AI and UX expertise. It is to use each where it creates the most value.
That may ultimately be the biggest change AI brings to the economics of UX. It doesn't make usability less important. It makes it harder to justify not doing anything about it.
