Around 1997 and 1998, I remember hearing a particular argument from photography enthusiasts and even some professionals that has stayed with me ever since. Digital photography was starting to become more visible, and there was a strong sense among some people that it simply wasn’t “real” photography. Film, in their minds, was where the skill was. A photographer had to understand exposure, light, film speed, development, lenses, timing and all the small technical decisions that went into making an image. Digital photography, by comparison, was often dismissed as something that would do too much of the work for you. The camera was becoming increasingly automatic, the computer was becoming more involved in the final result and, therefore, some believed the craft itself was being watered down.
There was also a very strong quality argument. I distinctly remember people saying that digital photography would never produce the same quality as film. At the time, that position was understandable. Early digital cameras had real limitations. Resolution was low compared with what people were used to from good film stock. Storage was limited. Sensors struggled in low light. Colours could look odd. Dynamic range was poor. Early digital images could sometimes look harsh or obviously electronic. If you were coming from years of carefully shot and processed film, it was easy to look at the first generations of digital cameras and conclude that this new technology was fundamentally inferior.
But that wasn’t really the point.
The point was that digital photography was new. It was developing quickly. And, more importantly, it was changing where the skill sat.
Today, virtually nobody would seriously argue that digital photography isn’t real photography. Modern digital cameras are astonishingly capable. They can identify eyes, track moving subjects, stabilise images, calculate exposure in fractions of a second, shoot at incredible frame rates and capture enormous amounts of visual information. Smartphones now use computational photography to combine multiple exposures, brighten shadows, reduce noise, sharpen details and correct colour before the photographer has even looked at the image.
Yet, despite all this automation, we still recognise the difference between an average photograph and a great one.
That difference still comes down to the photographer.
Technology may make the technical process easier, but it does not decide what is worth photographing. It does not automatically know the best place to stand, the moment to press the shutter, the expression to look for, the story that should be told or the feeling the image should create. A camera might calculate the exposure perfectly, but it cannot guarantee that the photograph will be meaningful. It can focus on the eye, but it cannot necessarily understand why that particular person, in that particular moment, matters.
Give two photographers the same camera, the same lens, the same location and the same subject and you can still end up with two completely different images. One might be technically fine but forgettable. The other might be striking, emotional or memorable. The tools are the same. The difference is judgement, experience, timing, composition, creativity and intent.
That is why I increasingly see a strong similarity between the transition from film to digital photography and the debate we are now having about artificial intelligence.
And photography is not the only comparison.
We Said Similar Things About Word Processors
There was also resistance when writing began moving from handwriting and typewriters onto computers.
In the late 1980s and through the 1990s, as word processors became increasingly common in homes, schools and workplaces, there were people who worried that computers would weaken basic writing skills. Features such as spellcheck were sometimes treated almost as a form of cheating. If a computer could identify a spelling mistake for you, then surely people would stop learning how to spell. If you could delete, move and rewrite whole sentences without crossing things out or starting a page again, perhaps writers would become less careful. If typing became easier than handwriting, perhaps we would eventually lose the ability to write properly with a pen at all.
I remember a version of this concern personally. My mother used to tell me that I should not forget how to use a pen because I was doing so much typing on a computer.
At the time, there was a certain logic to it. Writing by hand demanded a different sort of discipline. You had to think about what you were going to put on the page because correcting it was more difficult. A typewriter was only slightly more forgiving. A serious error could mean correction fluid, correction tape or, in some cases, retyping an entire page. A word processor changed all of that. Suddenly you could move paragraphs around, delete entire sections, rewrite a sentence ten times, copy and paste material and correct spelling errors almost instantly.
For some people, that looked like technology removing the skill from writing.
But that is not what happened.
Word processors did not destroy writing. They became the standard tool for writers.
Today, virtually nobody argues that an author has produced a less authentic piece of work because they typed it in Microsoft Word instead of writing it longhand. We do not dismiss a journalist because spellcheck caught a typo. We do not tell a copywriter that using Find and Replace somehow diminishes their ability. We do not demand that a novelist retype an entire manuscript simply to prove that editing requires enough effort.
Instead, we accepted that the technology removed some of the mechanical friction around writing.
And, importantly, it allowed the writer to focus more on the writing itself.
Spellcheck can point out that a word may be misspelled, but it cannot decide whether an argument is interesting. It cannot automatically make a dull paragraph compelling. It cannot understand every nuance of tone, context or meaning. It cannot decide whether a story deserves to be told.
The writer still has to do that.
In fact, the arrival of the word processor arguably made editing more important, not less important. Because it became easier to revise, writers gained the ability to restructure and refine their work far more freely. A paragraph no longer had to remain in the same place simply because moving it would require retyping three pages. An idea could be expanded, shortened or completely removed without starting over.
The tool changed the process.
It did not eliminate the craft.
And that is an important distinction when we talk about AI today.
AI Is the Next Version of the Same Debate
This week, a journalist said to me that AI needed to be stopped when it came to writing. The argument was essentially that AI-written articles were not of the same standard as articles written by humans and that allowing AI into publishing would ultimately reduce the quality of journalism and writing more broadly.
I understand the concern.
There is an enormous amount of poor AI content being created. Anyone who spends much time online has probably encountered it. There are articles full of repetitive sentences, generic observations, artificial enthusiasm and paragraphs that say a lot without really saying anything. There are websites publishing huge volumes of thin content simply because AI makes it cheap to produce. There are people using AI without fact-checking what it produces. There are organisations that appear to have confused the ability to generate words quickly with the ability to communicate something worthwhile.
But bad AI writing does not prove that AI cannot be used to create good writing.
There is plenty of terrible writing produced entirely by humans too.
The real question should not simply be whether a person typed every individual word manually. The more useful question is whether the finished article is accurate, thoughtful, interesting, useful and worth the reader’s time.
If a reader finishes an article feeling better informed, entertained, challenged or inspired, does it fundamentally matter whether AI played a role in helping to structure it, research it, edit it, improve the wording or develop some of the ideas?
I am not convinced that it does.
That does not mean the human role disappears. In many ways, I think the human role becomes more important.
AI can produce words, but somebody still needs to know what those words are supposed to achieve. Someone needs to understand the audience. Someone needs to know whether an argument is convincing. Someone needs to recognise when a piece of writing sounds bland or repetitive. Someone needs to identify when the AI has misunderstood the issue, invented a fact, overstated a claim or simply missed the point.
That judgement still sits with the person using the tool.
And when you put the development of AI alongside the rise of word processors, the resemblance becomes even clearer.
Spellcheck automated one part of writing.
Grammar tools automated another.
Search engines automated part of the research process.
Digital dictionaries removed the need to reach for a physical dictionary every time you were unsure about a word.
Cloud documents removed much of the friction involved in collaboration and version control.
AI simply goes further.
That is why it feels more disruptive.
But the fact that a tool goes further does not automatically mean that the underlying principle has changed. We are still using technology to reduce some of the mechanical effort involved in producing an outcome.
The question remains what the human does with the time, capability and leverage that technology creates.
The Same Thing Is Happening in Graphic Design
The same argument applies to graphic design.
For much of my career, producing a good visual could involve hours of hands-on work. You might sketch an idea, build it in Illustrator or Photoshop, refine individual shapes, adjust colours, move elements around, test different compositions and then spend another hour fixing small details that most people would never consciously notice.
Today, AI can sometimes create a useful image or diagram in a fraction of that time.
You can describe what you need, explain the subject, specify the angle, the style, the materials, the lighting, the colours and the intended purpose, and an AI system can generate a starting point almost instantly. You can then refine it, correct it, change proportions, alter materials, move elements and keep iterating until the image begins to match what you had in mind.
Some people look at that process and say there is no skill in it because the person did not physically draw every line themselves.
I don’t think that is a very useful way of looking at creativity.
The skill may simply have moved.
Instead of physically constructing every individual component, the creator is defining the outcome, directing the process, recognising problems and refining the work. In many ways, that starts to resemble creative direction more than traditional production.
And creative direction has always been a skill.
A creative director may not personally take every photograph used in a campaign. They may not draw every illustration, write every line of copy, animate every frame or build every webpage. Their role is often to understand the goal, determine what good looks like, brief specialists, evaluate the work and keep pushing the execution toward the desired result.
We do not normally dismiss that contribution simply because someone else physically pressed the camera shutter or moved the points in Illustrator.
AI introduces another tool into that process.
If anything, it puts more pressure on the person using it to know what they want.
A vague brief usually produces a vague result. A generic prompt often produces generic work. A person who has no understanding of design can ask AI for an image, but they may not notice that the proportions are wrong, the hierarchy is weak, the typography is poor, the composition is awkward or the image simply does not communicate what the business needs it to communicate.
The same is true in writing.
Someone with weak writing skills can generate an article quickly, but they may not notice that the argument goes nowhere. They may not recognise repetition. They may not know when a paragraph needs cutting, when an example needs expanding or when the tone is wrong for the audience.
AI can produce material.
It cannot always reliably determine whether that material is good.
That still requires judgement.
We Often Confuse Effort With Value
This is where I think some of the current conversation around AI becomes confused. We have a tendency to associate effort with value. If something took four hours to produce, we assume it must somehow be more valuable than something produced in 30 minutes.
But the audience rarely cares how long something took.
Imagine a company needs a technical diagram for a product brochure. One designer spends four hours manually illustrating it. Another designer uses an AI tool, creates an initial image in a few minutes, then spends another half-hour correcting and refining it until the result is accurate.
If the two finished diagrams are equally clear, equally professional and equally useful, what extra value did the four hours of manual labour necessarily create?
The customer does not normally look at the brochure and say, “I hope someone spent a long time making that.”
They look at whether the image helps them understand the product.
The same is true with writing. Readers do not value an article because somebody spent eight hours typing it. They value it because it told them something they wanted or needed to know.
The same principle is why we no longer think it is somehow more honourable to write a 2,000-word report by hand rather than type it. The manual effort involved has almost nothing to do with the quality of the thinking.
That distinction between effort and value has always mattered, but AI is forcing us to confront it more directly.
Technology has been reducing the amount of manual effort required to produce things for centuries.
We accepted calculators because they allowed us to perform calculations more efficiently.
We accepted spreadsheets because they allowed us to work with numbers at a scale that would have been painful by hand.
We accepted word processors because editing a document became easier than retyping pages on a typewriter.
We accepted spellcheck even though it technically meant the computer was helping us spell.
We accepted autofocus, automatic exposure and image stabilisation in cameras even though each of those technologies took over a task photographers once had to perform manually.
We accepted Photoshop, Illustrator, InDesign and computer-aided design even though they replaced many physical processes that once required considerable craft.
The arrival of AI is larger and more disruptive, but it is not completely different in principle.
It is another technology that changes the relationship between the creator and the process.
Maintaining Skills Still Matters
None of this means the foundational skills cease to matter.
My mother was right in one sense when she warned me not to forget how to use a pen. There is value in retaining the underlying capability even when technology makes one part of it easier.
I still know how to write by hand.
Writers should still know how to construct an argument.
Designers should still understand visual hierarchy, composition, colour and typography.
Photographers should still understand light, exposure and perspective.
Marketers should still understand audiences, positioning and commercial strategy.
Those skills are what allow us to judge whether technology is producing something worthwhile.
That is one of the most important points in the entire AI debate.
The danger is not simply that AI can do some of the work.
The greater danger is people relying on AI without possessing enough knowledge to recognise when the output is wrong.
Someone who understands a subject deeply can use AI as a powerful assistant. They can challenge it, refine it, correct it and reject weak ideas.
Someone who knows nothing about the subject may accept whatever the machine provides.
That is where quality can collapse.
The problem is not AI replacing expertise.
It is people trying to use AI as a substitute for having any expertise in the first place.
Technology Changes Where the Skill Sits
This is why I think the most useful way to understand AI is not as something that removes skill, but as something that changes where the skill sits.
With film photography, some of the skill was in handling film, judging exposure and developing images.
With digital photography, more of the process became automated and editable, but composition, timing, storytelling and judgement remained essential.
With handwritten or typewritten documents, part of the work involved the physical process of getting words onto a page and correcting them.
With word processors, that physical friction largely disappeared, allowing the writer to spend more time editing, reorganising and refining.
With AI, some of the initial generation can now be accelerated as well.
That means the human role increasingly shifts toward deciding what should be created, providing direction, evaluating the output, correcting problems and turning a generic result into something genuinely valuable.
That is not necessarily less skilled.
It is a different form of skill.
From Craftsperson to Director
Perhaps one of the biggest changes AI will bring to creative work is a shift in what being skilled actually means.
Historically, much of creative expertise involved execution.
Can you draw it?
Can you photograph it?
Can you write it?
Can you build it?
Increasingly, another question is becoming equally important:
Can you direct its creation?
Can you describe exactly what you need?
Can you distinguish an average result from an excellent one?
Can you provide meaningful feedback?
Can you combine ideas?
Can you identify inaccuracies?
Can you recognise opportunities the technology has missed?
Can you turn a rough AI output into something uniquely useful?
Those are not trivial abilities.
And as AI becomes easier for everyone to access, judgement may actually become more valuable.
If everybody has access to the same technology, the technology itself stops being the competitive advantage.
What you do with it becomes the advantage.
AI Will Not Make Everyone Equally Creative
Digital photography provides a useful lesson here.
Cameras became easier to use.
Photography became dramatically more accessible.
Billions of people now carry remarkably capable cameras in their pockets.
Yet exceptional photographers still exist.
Word processors did the same thing for writing.
Almost everybody can now open a document and produce perfectly formatted text. Spellcheck can correct obvious errors. Grammar tools can suggest improvements. Templates can make documents look presentable.
Yet that did not make everybody a great writer.
It simply reduced some of the technical barriers.
AI is likely to do the same thing on a larger scale.
It will lower the barrier to producing something that looks broadly competent.
That means there will be considerably more content.
Much of it will be average.
Some of it will be terrible.
But skilled people will also be able to use these tools to do things that previously required considerably more time, money or resources.
That is where AI becomes exciting.
A talented photographer with a modern digital camera can do things that would have been almost impossible with film.
A talented writer using a word processor can reorganise and polish a manuscript far more efficiently than someone working on a typewriter.
A talented designer using modern software can produce work that would once have required an entire production department.
And a talented person using AI can potentially explore ideas, produce drafts, create visuals and solve problems far more quickly than before.
The presence of the tool does not eliminate the value of the person.
It can amplify it.
The Better Question Is What We Do With the Tool
I understand why writers, designers, photographers and other creative professionals can feel uncomfortable about AI.
When a technology suddenly performs tasks that took years to learn, it is natural to question what that means for the profession.
Photographers faced a version of that question when digital cameras arrived.
Writers faced it when computers and word processors began replacing handwriting and typewriters.
Designers faced it as desktop publishing replaced traditional production methods.
Each time, something changed.
Certain manual processes became less important.
Some skills moved elsewhere.
New skills emerged.
But creativity continued.
AI is another useful tool.
It should not become an excuse to stop learning.
It should not remove responsibility for accuracy or quality.
It should not mean publishing whatever a machine produces without thought.
And it certainly should not mean that foundational skills no longer matter.
But neither should we assume that doing something manually automatically makes it more authentic, valuable or creative.
Sometimes the most skilled person is not the person doing the most manual work.
It is the person who understands the objective, chooses the right tools, makes the right decisions and produces the best result.
We accepted that lesson with word processors.
We accepted it with digital cameras.
I suspect, in time, we will accept it with AI too.
The pen did not disappear because of the keyboard.
The writer did not disappear because of spellcheck.
The photographer did not disappear because of autofocus.
The designer did not disappear because of Photoshop.
And AI does not automatically make creativity disappear either.
It changes the tools.
The human still has to decide what is worth creating.
