Over the past few years, AI has given a lot of people access to skills they didn’t have before.
If you struggled with writing before, you can now produce a passable piece of written content. If, like me, you have zero visual design skills, you can create an image (although the quality might be debatable). If you couldn’t code, these days you can build something that works.
That might feel incredibly liberating for you - a bit like a superpower. However, there are also people out there who have spent years developing those very skills you’re now able to turn over to a machine, and the experience of the last few years might feel very different for them. People like me - a writer who has been writing and reading for as long as I can remember. Someone who fell in love with words at an early age and treats writing like a calling, more than a job to do or something that makes me money.
I’ve been thinking about that difference a lot lately, particularly when I hear people say that it doesn’t matter if AI was used to create something, and that the only thing that matters is the output.
Here’s where I stand on this particular issue. AI didn’t acquire these capabilities from nowhere. They came from an enormous body of human knowledge and creative work - particularly people with specialist skills and subject matter expertise.
And if you ask many of us, it does matter where something comes from. It matters a lot.
Same technology, different experience
I think part of the disconnect comes down to the difference between what I might call specialists and generalists, particularly in the context of corporate or knowledge work.
A specialist has usually spent a significant chunk of their career, or even their life, developing deep expertise in a particular craft, discipline or knowledge area. A generalist works across a broader range of skills, connecting ideas and people without necessarily having the same depth in one particular area.
Most of us sit somewhere between the two - and these days, I probably do too, although I started life as more of a specialist.
My career has broadened considerably over the years. I’ve worked across media, corporate communications, PR, marketing and leadership. Today, I run a business and advise senior executives, which requires me to draw on generalist skills.
However, while I’ve been able to develop those skills across my career, writing remains at the core of who I am as a professional, if not as a human being. It’s the specialist skill I’ve had for most of my life, and long before it became my profession it was something I loved doing.
I know I’m not alone in this. For many specialists, their expertise isn’t simply something they learned because it offered good career prospects. It can be something they’ve been drawn to since childhood. We all know of the musician who plays when nobody is listening, or the designer who works on their art on the weekends. Or even the developer who builds things in their spare time because they genuinely enjoy solving problems.
How the onset of AI felt as a specialist
I would write whether I was paid for it or not, and that’s probably why I found some of the early commentary around generative AI surprisingly difficult. When ChatGPT arrived, I started hearing variations of the same idea: “now everyone’s a writer”, or even “who needs writers anymore?”.
I don’t think the people saying it intended to diminish writers. For someone who had always struggled with writing, it must have felt like a gift from above to suddenly have the ability to produce something they were happy with, without having to pay someone like me to write it for them.
I was anything but happy when ChatGPT arrived. It was devastating for me to see something I had spent most of my life developing, and which formed a meaningful part of how I saw myself, being discussed as a problem technology had solved.
I’m sure there’s plenty of specialists out there, just like me, who recognise this feeling. And many of you may have been wondering over the past few years where the empathy for us has disappeared to, as the thing we love doing gets increasingly handed over to robots - often with a sense of glee and “goodbye to those pesky specialists!”.
AI didn’t learn this by itself
Here’s the brutal truth I need to remind you of, if you’re one of the many people out there who likes to say “it doesn’t matter how the content was created”.
AI isn’t some magic wonder tool that appeared in the world fully formed. Generative AI is only capable of writing, coding, illustrating and answering highly specialised questions because it has been trained on enormous quantities of human-created material. It’s the end product of decades of accumulated knowledge created by people who became very good at what they do.
I’m deliberately not going to get into the legal arguments about how that material was collected and whether permission should have been required. Those questions are being debated elsewhere, and are still being fought through the courts.
The reminder I want to provide today is that AI’s capabilities came from somewhere. They came from people. People who spent a lot of time getting good at what they do, and are now being told in various ways that they no longer matter.
Provenance always matter, even with AI
This is the key reason I struggle with the argument that it doesn’t matter how something was created, as long as the output is decent. We don’t generally think that way about other things we consume - but somehow that’s become acceptable to some people in the context of AI.
We’ve become increasingly conscious of where our clothes are made and the conditions of the people who make them. We want to know where our food comes from. We care whether products have supply chains we’re comfortable supporting. The finished product doesn’t render those questions irrelevant.
While I’m not suggesting the ethical issues surrounding AI training are equivalent to exploitation in global supply chains, I am making a point about provenance. Where something comes from still matters, even when we’re perfectly happy with the thing we receive at the end.
And if you’re a specialist, the phrase “it doesn’t matter where it came from” can land particularly badly - because “what it came from” was people like you.
So when someone says the provenance doesn’t matter because the output is good enough, what they may intend as an empowering observation can sound very different from the other side. It can sound like the original contribution doesn’t matter either.
Am I asking you to stop using AI?
Believe it or not, no. I’m not asking anyone to stop using AI because some specialists are uncomfortable with what’s happening. If AI has given you the ability to do something you previously struggled with, I can understand why you’re excited about it.
What I’m asking for is a little empathy for the people experiencing the same technology from a very different position.
If you’re a project manager who has always struggled to write, having AI help you produce a report might feel like a superpower. If you’re running a small business and can suddenly create your own marketing materials, that might save you money you simply don’t have.
For someone who has spent 30 years becoming a writer, illustrator or developer, things might feel a little more complicated. People like me are watching technology become increasingly capable of producing the kind of work they’ve spent much of their lives learning how to do. At the same time, we’re hearing people celebrate the fact that specialist expertise is becoming easier to access.
You can be excited about what AI makes possible for you while recognising that someone else might feel differently about what it means for them.
If it doesn’t matter, why not disclose it?
This brings me to something I’ve been thinking about more seriously: disclosure. If AI played a significant role in creating something you’ve produced, I think you should consider saying so.
I’m not talking about using AI to check your spelling, help with research or suggest a better way to structure a sentence. I mean using generative AI to substantively create the writing, artwork, code or other work that you’re presenting. If it doesn’t matter where it came from, why not tell people?
There’s no accusation embedded in that question. Using AI isn’t inherently dishonest and disclosure doesn’t need to be an admission that you’ve done something wrong. It’s simply transparency about how the work was produced.
We already accept this principle elsewhere. Labels tell us where products were manufactured. Businesses increasingly provide information about their supply chains because consumers want to understand more than what they see in the finished product.
I think there’s a reasonable case for bringing some of that thinking into AI. Disclosure acknowledges that the process matters alongside the output. It recognises that these capabilities have a provenance, and that human expertise sits behind them. It’s also a small way of showing some consideration for the specialists whose work helped make your current opportunities possible.
If your position is genuinely that it doesn’t matter how something was created because only the output matters, disclosure shouldn’t be a problem. On the other hand, if the idea of telling people you used AI makes you uncomfortable, it could be worth asking yourself why that is.
Perhaps where the work came from matters more than some of us are willing to admit.
One thing I’m thinking about this week
While I was writing this piece, I found myself wondering what’s next.
AI has made specialist capabilities much more accessible, but what happens if that also makes it harder for people to earn a living developing those capabilities and knowledge in the first place? What if the gifted artist no longer makes a living from designing, or the writer from writing, and so on?
There’s some fascinating research emerging around AI’s continued dependence on fresh human-generated knowledge, and the risks of models increasingly learning from synthetic material.
This raises the question I want to explore next. If AI still needs human expertise, what happens if we stop giving people a reason to develop it?
Until next week… I’ve got some research to do!
Things I’ve made (and you can buy)
If you enjoy my content and would like to support my work, here are a few things I’ve created.
🛒 Track Changes On Shop - Coffee mugs and drinkware for corporate rebels
📖 Do Give Up Your Day Job - Guide to corporate exits and transitions to self-employment
