Am I Worried About the AI Boom? Not Really.

There is probably an obvious question to ask somebody who makes websites for a living at the moment:

Am I worried about AI?

A little, obviously.

It would be strange not to pay attention to technology that can now generate a functioning website, write the copy, create images and build surprisingly complicated applications from a conversation.

But am I sitting here thinking web designers and developers are about to become irrelevant?

Not really.

If anything, the last year has made me slightly less worried.

And a big part of the reason is that people are already getting tired of AI.

Not AI itself.

AI slop.

There is a difference between AI and AI slop

I use AI.

Quite a lot, actually.

It can speed up development, help explore an idea, find problems in code, generate a starting point, research something unfamiliar or get through repetitive work much faster than I could manually.

Pretending none of that is useful would be ridiculous.

But there is a very noticeable difference between using AI as part of the process and simply publishing whatever it gives you.

You can see the latter everywhere now.

The same illustrations.

The same glowing gradients.

The same rounded cards.

The same oversized headings.

The same strangely enthusiastic copy telling you how a company is going to “revolutionise” something.

The same em dashes.

The same people with six fingers if somebody forgot to look closely enough.

Eventually you stop noticing that something was made with AI and start noticing that nobody really finished it.

That distinction matters.

People are already learning what AI looks like

What has surprised me is how quickly the resistance has appeared.

“AI slop” has become a fairly normal phrase.

People actively complain about generated imagery, generated articles, generated adverts and websites that feel like nobody made any decisions beyond accepting the first plausible output.

There are now entire conversations in design communities about whether a website looks “too AI”. Some designers are even finding work they made manually being accused of using AI simply because it shares the visual patterns that AI generators have heavily reproduced.

That is quite an interesting reversal.

For years, digital design moved towards cleaner grids, smooth gradients, perfect illustrations and increasingly polished interfaces.

Now being too generically polished can itself become suspicious.

I wrote something related to this before in Why Imperfect Web Design Often Feels Better.

Sometimes the little irregularities are exactly what tell you somebody was there.

The barrier to making something has collapsed

This is the part of the AI boom that I do think is genuinely enormous.

The barrier to making things has dropped dramatically.

Someone who has never written code can now build a working prototype.

Someone who could never afford a custom site can get a decent-looking website online.

A founder can test an idea without paying thousands before knowing whether anybody wants it.

A small business owner can write a first version of their own content instead of staring at an empty page.

I think all of that is good.

I do not think protecting web design means wishing those tools did not exist.

The interesting question is what happens after everyone can make something.

And I think we are starting to find out.

When everybody can generate, generation becomes less valuable

If it takes thirty seconds to generate a homepage, having a homepage is no longer particularly impressive.

If everyone can generate ten logo ideas, generating a logo idea is not where the value is.

If an AI can produce twenty headlines instantly, producing headlines is no longer the difficult part either.

The valuable bit shifts.

Which headline is actually right?

Which section should disappear?

Does this business even need that feature?

Why does the mobile version feel awkward?

Does the photograph feel believable?

Does this sound like the person behind the company?

Is this website genuinely easier to use or does it just contain more stuff?

At what point is it finished?

Those are judgement questions.

AI can give you answers to them too, of course.

But eventually somebody still has to decide whether those answers are any good.

That is something I touched on in What Actually Makes a Website Feel Professional?.

A lot of professional design is not producing more.

It is noticing small things and making consistent decisions about them.

AI is incredibly good at adding things

This is probably one of my biggest frustrations with AI-assisted design.

Ask it to make a page better and its instinct is usually to give you something else.

Add some cards.

Add a gradient.

Add a badge.

Add three statistics.

Add an animation.

Add an icon row.

Add a section explaining the thing the previous section already explained.

It feels productive because something visibly changed.

I am increasingly convinced that good website design often works in exactly the opposite direction.

You remove the redundant section.

You shorten the introduction.

You give the photograph more space.

You decide that one button is enough.

You take away the animation because it is distracting from the thing somebody actually needs to read.

I wrote an entire article about this recently: I’ve Realised Better Websites Don’t Always Need More.

AI has actually reinforced that opinion for me.

Generation is becoming cheap.

Editing is becoming more important.

Generic used to mean safe. Now it can mean AI.

There is another change happening which I find particularly interesting.

For a long time, businesses often wanted websites that looked broadly like other successful websites.

Clean.

Professional.

Modern.

Safe.

There was not necessarily anything wrong with that, but it produced a lot of websites that were technically good and completely forgettable.

AI has accelerated that look so aggressively that being generic may actually start carrying a cost.

If a visitor has already seen fifty websites with the same gradient hero, floating dashboard mockup and three glowing feature cards, the fifty-first one does not feel futuristic.

It feels generated.

That could create a strange incentive for businesses to become more individual again.

Real photography.

More distinctive typography.

Less perfect layouts.

Specific copy.

Odd details.

A stronger point of view.

Maybe even a little bit of mess.

That seems like a fairly healthy direction to me.

There is a similar problem with the words

The design is only half of it.

Generated copy has its own fingerprint.

There is nothing inherently wrong with using AI to write.

I use it when it helps.

But there is a certain kind of AI-written website where every paragraph feels perfectly grammatical and somehow says absolutely nothing.

A business is “passionate about delivering tailored solutions.”

It offers a “seamless experience.”

Its team “leverages cutting-edge technology.”

Everything is apparently going to be elevated.

Nobody talks like this.

Or perhaps more accurately, nobody talks like this until they ask an AI to write their website.

The irony is that as generated writing gets technically better, normal language becomes more noticeable.

A slightly blunt sentence can feel refreshing.

A specific opinion can feel trustworthy.

A business owner actually explaining why they do something a certain way can suddenly stand out.

That does not feel like a terrible future for writing.

I don't think people care whether AI was involved

This is an important distinction.

I do not think most customers ultimately care whether somebody used AI while making their website.

They probably should not.

Clients do not ask whether I used Stack Overflow.

They do not care which autocomplete tool helped write a function.

They do not care whether Photoshop used content-aware fill.

They care about the result.

The issue starts when the tool becomes visible in the result.

If the website feels generic, inaccurate, unfinished or strangely disconnected from the business, then how it was produced suddenly becomes relevant.

But if AI helped somebody make something genuinely good, I doubt most people will object simply because AI was involved somewhere in the workflow.

I think the backlash is less anti-AI than it sometimes appears.

It is anti-low-effort.

AI website builders are going to get much better

I also do not want to base my confidence on today's limitations.

AI builders will improve.

The code will get better.

The designs will become less repetitive.

Agents will get better at testing what they build.

They will understand existing websites more deeply.

They will make fewer strange decisions.

Things that currently need a developer will increasingly become possible through conversation.

I have written before about AI website builders compared with custom web design, and I would probably write parts of that article differently a year from now.

The tools are moving too quickly not to.

But that does not necessarily lead me to the conclusion that there will be no work left.

It changes where the work sits.

I suspect there will be a lot more unfinished websites

One thing I am already seeing is people getting much further by themselves.

That creates a slightly different kind of client.

They are not coming to a developer saying:

“I have an idea for a website. Can you build it?”

They are saying:

“I've built most of this. Can you make it actually good?”

That is a completely different starting point.

Maybe the site needs better design.

Maybe the code needs reviewing.

Maybe the SEO structure makes no sense.

Maybe everything works except mobile.

Maybe an AI helped build an impressive-looking SaaS prototype but nobody has really looked at authentication, permissions or what happens when something fails.

This is partly why I recently added an AI Website Redesign & Cleanup service.

I do not see AI-built websites purely as competition.

Some of them are becoming the raw material.

The same thing happened with website builders

None of this is entirely new.

Wix did not kill web design.

Squarespace did not kill web design.

WordPress themes did not kill web design.

Elementor did not kill web design.

Shopify did not kill ecommerce development.

They changed who could build a website and what people were willing to pay somebody else to do.

AI is obviously a much bigger shift than another website builder, so I do not think those comparisons are perfect.

But the pattern is familiar.

Tools remove some work.

Expectations then move upwards.

Twenty years ago, simply having a website was impressive.

Then having a responsive website became normal.

Then businesses expected online booking, ecommerce, animation, integrations, dashboards, SEO, analytics and content management.

The baseline keeps moving.

AI will move it again.

Some parts of my job probably will disappear

I also think it would be dishonest to pretend there is no threat at all.

There is work people currently pay for that AI will make difficult to charge for.

Very simple brochure websites will become easier.

Basic copywriting will become cheaper.

Routine development tasks will take less time.

Simple logo ideation will be commoditised further.

A lot of technical knowledge that used to require hiring someone will become accessible through a conversation.

That is significant.

If somebody's entire service is essentially assembling standard components into a standard website, I think the next few years could be uncomfortable.

But I am not sure preserving that work indefinitely should be the goal.

The bit I want to compete on is taste

“Creative judgement” sounds slightly pretentious.

Taste is probably the simpler word.

Knowing when something feels wrong.

Knowing which reference is actually relevant.

Knowing when a design has gone too far.

Knowing when the technically impressive option is worse for the person using it.

Knowing when a page should be longer.

Knowing when it should be half the length.

Knowing when the client is asking for something that will genuinely improve the project and when they are reacting to a temporary preference.

None of that means a human is magically always right.

We are very capable of making terrible decisions ourselves.

But it is the part of the job I find most interesting, and AI making production easier gives me more reason to focus on it rather than less.

Maybe the boom makes human work more obvious

There is an irony here.

The more generated content fills the internet, the easier genuinely specific work can become to notice.

A real photograph stands out.

A strange little interaction stands out.

Copy that sounds like someone actually wrote it stands out.

An opinion stands out.

A website that does not look like every other startup homepage stands out.

That does not require rejecting technology.

It requires using it without allowing it to flatten everything into the same answer.

This connects quite closely with something I wrote in Are We Building Websites for Algorithms Instead of People?.

Different technology, similar danger.

When you optimise too aggressively for a system, it is very easy to forget who the thing is actually for.

So, am I worried?

Enough to keep learning.

Enough to change what I offer.

Enough to accept that parts of this job will look very different in a few years.

But not enough to think good designers and developers suddenly have no value.

AI is making it easier to create.

It is also flooding the internet with more things that technically exist but do not feel considered.

And people seem to be noticing the difference surprisingly quickly.

Maybe the future of web design involves far more AI than it does today.

I expect it will.

But I also suspect the businesses that stand out will be the ones where somebody still cared enough to edit the result, question it, remove things, make unusual decisions and occasionally ignore what the machine suggested.

If that is where things are heading, I am quite comfortable with it.