AI Didn't Replace the Designer. It Replaced the Bad Poster.

There is an assumption sitting underneath a lot of the conversation about whether AI will replace creative jobs.

You see an AI-generated poster and imagine a graphic designer who didn't get paid.

You see an AI-written website and imagine a copywriter who lost the job.

You see a small business using an AI-generated logo and imagine an illustrator whose commission disappeared.

Sometimes that is exactly what happened.

AI is replacing some paid professional work, and pretending otherwise would make any defence of the technology considerably weaker.

But I suspect we're applying that explanation to far more AI-generated content than it actually fits.

Because there is another possibility.

The professional was never getting that job in the first place.

Will AI replace creative jobs, or are we creating things that never existed before?

There is unquestionably an explosion of AI-generated stuff.

Posters. Blogs. Social media posts. Websites. Logos. Videos. Newsletters. Product descriptions. Flyers. Presentations.

A great deal of it isn't particularly good.

It gets labelled "AI slop", sometimes fairly.

But there is a mistake hidden inside the observation.

We see more poor AI content and assume that poor AI content has replaced good professional content.

What if much of it has actually replaced no content at all, or the equally poor content people were already making themselves?

That local football club was not previously paying an agency to design its summer tournament poster.

The village charity was not employing an advertising firm for its Christmas raffle.

The new window cleaner was not commissioning a copywriter, photographer, branding consultant and web developer before putting up a basic website.

The amateur theatre group did not have an annual creative budget.

They had volunteers.

They volunteers had Microsoft Word.

The volunteers were doing their best.

AI has not necessarily taken those commissions away from creative professionals because, in many cases, those commissions never existed.

We need to compare AI with what would actually have happened

This is the bit I think gets missed.

When somebody creates something with AI, we instinctively compare it with the best professional alternative.

AI poster versus professional graphic designer.

AI copy versus professional copywriter.

AI website versus professional web developer.

But that isn't necessarily the real choice the person was making.

Economists would call this the counterfactual. In plain English, it means asking:

What would actually have happened if AI wasn't available?

Would the business have hired a professional?

Sometimes, yes.

In that case there is a reasonable argument that AI has displaced paid creative work.

But perhaps the owner would have made it themselves.

Perhaps a volunteer would have done it.

Perhaps an intern would have been given the job.

Perhaps somebody's nephew "who is good with computers" would have had a go.

Perhaps they would have used a template.

Perhaps nothing would have been produced at all.

Those possibilities matter.

Because the alternative to AI isn't automatically a talented professional being properly paid.

Sometimes the alternative is a shit poster.

I know because I used to be the cheap alternative

Before I became a graphic designer

I designed some posters and Christmas cards for a local charity.

They were awful.

Not charmingly amateur.

Just awful.

But the charity didn't choose me instead of an experienced professional designer charging proper commercial rates.

There wasn't a board meeting where somebody compared my portfolio with three established studios and decided to save the money.

They used me because I was available, willing and free of charge.

The realistic alternative was probably somebody else making something equally bad.

Or nothing being made at all.

That's an important distinction when we talk about AI replacing creative work.

There has always been an enormous layer of creative production taking place beneath the professional market.

Volunteers did it.

Students did it.

Interns did it.

Friends and relatives did it.

Someone in the office who "knows Photoshop" did it.

The owner stayed up until midnight doing it themselves.

And very often, it looked exactly like you would expect work made by someone with no training, no budget and no time to look.

AI has entered that market too.

Possibly more aggressively than it has entered the professional one.

The terrible AI poster might actually be an improvement

Imagine a small community group needs a poster.

Ten years ago they might have opened Word, centred some Comic Sans, stretched a JPEG they found online, added six different colours and printed 40 copies.

A few years ago they might have opened Canva and picked a template.

Today they might ask an AI tool to generate some artwork, drop it into a template and add the event details.

The result may still make a professional designer wince.

There might be too much text.

The spacing might be wrong.

The hierarchy might be questionable.

The AI-generated person in the corner might have developed an ambitious number of fingers.

It is not professional design.

But compared with what this particular organisation would realistically have produced before AI, it might be substantially better.

So whose job has been lost?

Potentially nobody's.

If we want to understand the effect of AI on creative work properly, we have to compare the output with the real alternative, not the professional output we wish the organisation had commissioned.

The same thing is happening with writing

Before generative AI, a small business owner who couldn't afford a copywriter didn't simply leave the internet out of respect for the profession.

They wrote the copy themselves.

You've read it.

"Welcome to our website. We are a family-run business offering a friendly and professional service. No job too big or too small."

The spelling might have been questionable.

The service page might have been three sentences long.

Nobody had thought about search intent, structure or what a potential customer actually wanted to know.

That business didn't have £3,000 sitting in an envelope marked COPYWRITER and suddenly redirect it towards ChatGPT.

There was no £3,000.

Now the owner can explain what the company does in their own words, give that information to an AI tool and ask it to help organise or improve the writing.

Is the result necessarily as good as an experienced copywriter who has researched the audience, interviewed the business and spent years learning their craft?

Of course not.

Might it communicate the business considerably better than what would otherwise have existed?

Absolutely.

Again, that isn't necessarily professional work being replaced.

It can be amateur work becoming less amateur.

AI has made competence cheaper

That might be one of the most consequential things generative AI has done.

Not excellence.

Competence.

Those are different things.

Expertise still exists.

Taste still exists.

Judgement still exists.

Experience still exists.

A brilliant designer can still do things with hierarchy, communication and visual ideas that somebody generating their first graphic cannot.

An experienced writer understands audience, rhythm, persuasion, research and context in ways that aren't reduced to grammatically correct sentences.

A good developer understands architecture, security, maintainability and all the things that happen after somebody has managed to make a prototype work once.

But the floor has moved.

It is now considerably easier for somebody with very little money or training to produce something acceptable.

That matters.

For decades, polished communication required some combination of skill, money and access.

If you lacked the skill, you paid somebody.

If you lacked the money, you learned it yourself.

If you lacked the time to learn it as well, you settled for something worse.

AI adds another option.

And it's available to people who previously had remarkably few.

Calling AI use "lazy" starts to become uncomfortable

This is where some of the criticism becomes more interesting.

A common response to AI-generated work is:

"Stop being lazy and do it yourself."

Or:

"Pay a professional."

But think about what those two options actually mean.

Do it yourself, even if you lack the skill, time, confidence or ability.

Or pay somebody, even if you cannot afford them.

Those are very easy options to prescribe when you personally have either the skills or the money.

A plumber starting a business might need:

• A logo

• A website

• Business cards

• Flyers

• Social media graphics

• Customer emails

• Quote templates

• Photographs

• Advertising copy

• Terms and policies

Are we seriously saying the morally correct route is for that plumber to learn branding, graphic design, photography, web development and copywriting?

Otherwise they should employ six different professionals?

They've got plumbing to do.

Before AI, they probably didn't hire all those people.

They muddled through.

They used a website template.

They asked a relative to make the logo.

They took photographs on their phone.

They wrote the homepage themselves after work.

They copied and pasted the same Facebook post into three different groups.

AI hasn't necessarily removed six professional commissions.

It may simply have improved the self-service layer that was already there.

Rich people have always outsourced

This is where calling AI users lazy can start to look remarkably selective.

Wealthy people and wealthy organisations have always outsourced work.

They employ:

• Assistants

• Designers

• Copywriters

• Developers

• Accountants

• Researchers

• Consultants

• Photographers

• Lawyers

• Marketing agencies

• Entire communications departments

Nobody looks at a managing director presenting a beautifully designed deck created by six employees and says:

"Lazy. You should have designed those slides yourself."

An executive can ask an assistant to summarise a 30-page document.

That's delegation.

A business owner can pay a copywriter to rewrite an email.

That's professional support.

A company can pay an agency to write 100 product descriptions.

That's business.

A politician can deliver a speech written by somebody else and nobody is confused about how professional speechwriting works.

We have never consistently believed that everybody should personally perform every task associated with their work.

We accept specialisation.

We accept delegation.

We accept outsourcing.

What AI changed is that a form of outsourcing became dramatically cheaper.

And suddenly we became intensely interested in whether the person "did the work themselves".

AI didn't invent outsourcing. It democratised some of it.

If somebody with money pays a human assistant to summarise a report, we call it delegation.

If somebody without an assistant asks AI to summarise it, we sometimes call it laziness.

If a large company hires an agency to produce hundreds of pieces of marketing content, that's an investment.

If a sole trader uses AI because they cannot afford the agency, they're accused of cutting corners.

Neither person necessarily did the work themselves.

The difference is that one outsourced through capital and the other outsourced through technology.

That doesn't automatically make every use of AI good.

But it does make the accusation of laziness considerably more complicated.

AI didn't invent outsourcing. It made some forms of outsourcing available to people who could never previously afford them.

Calling that laziness risks turning privilege into virtue.

When doing the work yourself actually matters

There are situations where this argument doesn't apply.

If a student is being assessed on their writing ability, their own work matters.

If an artist sells something specifically as hand-drawn when it was generated by AI, the deception matters.

If a journalist claims to have conducted reporting that never happened, that's a serious problem.

If somebody uses AI to fabricate expertise, evidence or experiences they don't have, that's a problem.

If the purpose of the activity is specifically to assess or purchase the person's own skill, secretly substituting AI changes what is being offered.

The question shouldn't therefore be:

"Did they use AI?"

A better question is:

"Was this a situation where their personal authorship, labour or skill was actually part of what was being promised?"

For an examined essay, yes.

For a commissioned original illustration sold as the artist's own work, yes.

For a bowls club poster advertising Saturday's fundraiser?

Probably not.

AI can give people access they didn't previously have

The argument also extends beyond money.

The person who struggles to write confidently can now get help expressing an idea.

Someone with dyslexia can turn scattered notes into structured writing.

A person who speaks English as an additional language can communicate professionally without their grammar becoming the first thing a prospective customer judges.

Someone who cannot code can prototype an idea.

A tiny club can produce newsletters.

A local campaign can create graphics.

A sole trader can build a basic website.

A parent can make invitations for their child's party without spending an evening fighting Microsoft Publisher.

The quantity of content increases because the number of people capable of producing it increases.

That isn't unprecedented.

Every democratisation of creativity produced a lot of rubbish

Cheap digital cameras created millions of terrible photographs.

WordPress created millions of dreadful websites.

YouTube created an extraordinary quantity of bad video.

Canva created an ocean of questionable graphic design.

Desktop publishing allowed practically anyone to create a leaflet, and practically anyone did.

Social media allowed everyone to become a publisher.

They published.

Giving millions of people access to a tool does not produce millions of masterpieces.

It produces millions of people using the tool.

AI is doing the same thing at extraordinary speed and across several disciplines simultaneously.

The fact that much of the output is bad doesn't, by itself, tell us that expanding access was a mistake.

But the criticism of "AI slop" isn't entirely wrong

There is a genuine problem here.

If everyone can create content, everyone creates content.

The cost of generating another article, illustration, social post, song, video or presentation can approach zero.

So we get more.

Far more.

Good things become easier to make.

So do pointless things.

Before generative AI, producing something required enough effort that the effort itself acted as a crude filter.

It wasn't a fair filter. People with money could simply pay somebody else.

But it was still a filter.

AI dramatically reduces that friction.

The internet doesn't merely become more democratic.

It becomes noisier.

We may therefore be moving from a problem of access to creation towards a problem of attention in a world of unlimited creation.

That's a legitimate concern.

It still doesn't follow that the people newly able to create things should never have been given the tools.

It means abundance creates different problems from scarcity.

And yes, professional creative jobs really are being affected

None of this means professional creatives are safe.

They aren't.

Some businesses that previously paid for professional work are now using AI instead.

Some companies will decide that acceptable work produced cheaply is preferable to excellent work produced professionally.

Some already have.

Entry-level work may be particularly vulnerable because many of the tasks traditionally handed to junior employees are precisely the sort of tasks AI can perform reasonably well.

That creates another problem which deserves much more attention.

Today's experienced designer was once an inexperienced designer.

I certainly was.

Those dreadful charity posters were part of how I learned.

Junior writers become senior writers by writing.

Junior developers become experienced developers by solving problems.

If AI removes a significant amount of the beginner-level work through which people traditionally develop expertise, where do tomorrow's experts acquire their experience?

That's a far harder and more important question than simply saying:

"AI stole someone's poster job."

AI can simultaneously expand access to creative capability and damage the career ladder that produces future professionals.

Both things can be true.

Large businesses haven't suddenly stopped needing expertise

There is also a danger of looking at millions of AI-generated social posts and assuming expertise itself has disappeared.

Large organisations still employ designers.

They still employ developers, marketers, writers, photographers, videographers and agencies.

Those professionals are increasingly using AI too.

There certainly are companies reducing headcount through automation, and the long-term employment effects remain uncertain.

But we should distinguish between:

A company replacing paid professional work with AI

and:

Someone who never had access to paid professional work suddenly being able to produce something themselves.

Both result in more AI-generated content.

Only one necessarily represents a lost commission.

From the outside, they can look identical.

Professional creative work may actually become easier to recognise

There is an irony here.

When basic competence becomes cheap, competence itself stops being particularly impressive.

A clean layout is no longer enough.

Grammatically correct copy is no longer enough.

A functioning brochure website is no longer enough.

A plausible illustration is no longer enough.

That's uncomfortable for professionals whose commercial value depended partly on being able to execute those things.

But it may push genuine professional value somewhere more interesting.

Taste.

Judgement.

Ideas.

Understanding an audience.

Knowing what to leave out.

Knowing when the obvious answer is wrong.

Knowing why something doesn't work even when the client cannot explain what feels wrong.

Understanding a business rather than merely producing an asset for it.

Creating something that doesn't resemble the statistical average of everything that came before it.

AI may not remove the need for experts.

It may remove some of the artificial scarcity that allowed basic execution to masquerade as expertise.

AI hasn't completely levelled the playing field

It would be equally simplistic to finish by declaring that AI has democratised everything and we're all equal now.

We aren't.

Money still matters.

Education matters.

Taste matters.

Connections matter.

Time matters.

Access to technology matters.

The best AI systems themselves are controlled by enormously powerful companies with enormous resources.

AI can democratise creative capability at one end while concentrating technological power at the other.

There are also legitimate unresolved questions around copyright, training data, environmental impact, misinformation, employment and deception.

Those aren't objections we should wave away because AI makes it easier for a local club to design a poster.

Technology can be useful and disruptive at the same time.

The comparison we should actually be making

The next time you see an AI-generated poster, website, illustration or piece of copy, don't automatically compare it with what an excellent professional could have produced.

Ask what would actually have existed without AI.

Would they have hired the designer?

If yes, perhaps AI displaced professional work.

Would the owner have made it themselves badly at midnight?

Would a volunteer have done it?

Would an intern have been given the job?

Would they have used a template?

Would nothing have been made at all?

Those possibilities matter.

Because if the real alternative was poor output or no output, AI hasn't necessarily replaced a creative professional.

It has allowed somebody who was never really participating in the professional market to make something better than they could before.

That is a very different story.

And perhaps that's why the AI debate becomes so heated.

The same technology can simultaneously threaten one person's advantage and remove another person's disadvantage.

For a professional creative, cheaper access to competent design, writing or development can feel like devaluation.

For the person who could never afford those services in the first place, it can feel like liberation.

Both experiences are real.

The mistake is pretending only one of them counts.