OpenAi’s new image generator may not just democratise design, but upend creative hierarchies and entire industries.
This week I’ve been playing with something that feels like a moment – one of those quiet milestones you look back on and realise everything changed. OpenAI’s new image generator, now built into ChatGPT, is extraordinary. You give it a prompt and within seconds, it produces not just an image, but a beautifully composed piece of design – layout, fonts, colours, even the copy. No templates, no training, no fiddly software. Just elegant output from a single sentence.
What makes it different from earlier versions is its almost eerie polish. The spelling is faultless. The proportions are balanced. The typography feels hand-picked by a design graduate. You can create lifelike portraits, infographics, marketing posters, packaging concepts – anything for which you might once have paid a creative agency thousands. It’s as if a slick Clerkenwell design team has taken up residence in your back pocket. Except they’re free.
Naturally, this is causing waves. Graphic designers, understandably, are worried. Photographers too.
For years, people assumed photography was one of the last creative disciplines AI would struggle to replicate convincingly. You can’t fake the energy of real people at a party or the atmosphere of a studio shot, they said. But this new model comes alarmingly close. And while it may not yet replace the magic of the perfect shot, it does something potentially more disruptive: it removes the cost barrier. What was once the preserve of global brands and glossy magazines can now be done by anyone with a phone and an idea. You can create stunning campaign visuals for your village fête, your pop-up bakery, your local bookshop, without needing to know a soul in Soho or spend a penny on a photographer.
We’ve been here before, of course. Technological disruption always starts like this – as a toy, then a tool, then a threat. When the printing press was invented in the 15th century, professional scribes across Europe protested. Their livelihood – copying manuscripts by hand – vanished almost overnight. The same resistance greeted the arrival of newspapers, and again when they went digital. People said no one would read the news on a screen. They were wrong.
When television arrived, it was supposedly going to destroy books. In fact, it created a new kind of shared experience. When I was growing up, television meant three black-and-white channels, one of which required careful dial-turning to find. And yet it brought families together. Entire neighbourhoods would sit down separately but simultaneously to watch the same programme. It created a kind of unspoken community – a shared story.
The internet was supposed to ruin that. Instead, it allowed people to find each other across distance and difference. WhatsApp groups and online forums now organise camping trips, school fêtes, neighbourhood watch schemes and rare dog rescue missions. The way we connect changed, but the desire to connect didn’t.
The arrival of this image generator feels like another one of those moments – a shift so subtle that many will miss it, yet so profound that it will change everything. And it matters, not just because it’s clever, but because of who it might empower.
I didn’t go to private school. I’ve never had the luxury of backing from big money or inside connections. I now lecture on entrepreneurship at Cambridge, and I’m aware, every time I walk into that room, that many of the people I speak to come from backgrounds where finance is a given, not a hurdle. They’ll get investment for ideas that wouldn’t last five minutes in the real world, simply because of who they are, where they studied and who they know. Meanwhile, great ideas from talented but neurodivergent or working-class founders often never get off the ground. The odds are stacked.
The numbers say it all. Around 7% of the UK population attended private school. Yet over 40% of FTSE 100 CEOs did. Nearly two-thirds of senior judges and senior civil servants come from the same tiny pool. The Sutton Trust has shown repeatedly how Oxbridge graduates continue to dominate Britain’s power structures – in politics, the media, the City. It isn’t a level playing field. But something like this new wave of AI could start to change that.
Because now, anyone with a smartphone can design a brand, produce a marketing campaign, check the legal and accounting position, write a business plan and create their own visual identity – without gatekeepers. A single person, working hard in their bedroom or café, can do what once took a team of professionals.
They still need to connect with people. They still need to pitch, build, raise money and execute. But they don’t need the full agency and advisory entourage from day one. The first hurdles are suddenly lower.
I’ve watched the excitement around AI rise and fall over the last year, and I think much of the hype has missed the point. Too much value has been placed on owning an AI business. But the real winners won’t necessarily be those who build the tools. They’ll be the ones who use them smartly. Nvidia and OpenAI have soared in value because they make the infrastructure. But companies like Amazon, which embed AI into everything from logistics to content, will be the quiet titans. This technology is becoming invisible – like electricity. We won’t even notice it, but it will be doing everything.
And then there’s China. A lot has been said about its tech race with the West, but I don’t think people fully grasp the potential. This isn’t just about Chinese companies. It’s about 1.3 billion highly motivated, increasingly entrepreneurial individuals who no longer just copy the West but are learning to lead. They’re unencumbered by legacy systems or nostalgia for how things used to be. I believe that within three years, we’ll see China become the most prolific and dominant financial engine the world has ever known – powered not by factories, but by ideas.
Closer to home, I work with traditional publishing companies, trying to help them modernise. Our business uses AI to help them produce more content, more affordably, so they can spend more time on real journalism and make their business models sustainable again. I’m not here to cut jobs. I’m trying to stop these businesses from folding.
The sad truth is that while many legacy organisations remain paralysed by caution, younger disruptors – and entire countries – are using these tools already, without hesitation. And they will not wait.
Of course, this all comes with risk. Deepfakes, disinformation, visual manipulation – we’re entering an age where seeing is no longer believing. But what’s interesting is how quickly we’re already adapting. There’s a move towards a more lo-fi aesthetic – raw, unfiltered, iPhone-style imagery that feels real, even if it’s less “perfect.” Magazines and news organisations are already leaning into it – roughness as a badge of authenticity. The digital Polaroid.
In the end, I believe humans will always find a way. We adapt. We recalibrate. We repurpose. AI is just another tool – one with enormous potential, yes, but still just a tool. The advantage will always lie with those who are willing to learn, experiment and make something new. And with the right mindset, this new generation of tools won’t replace us – they’ll simply help us do what we were already doing, but better.
Ivan Massow is a co-founder of Tomorrow’s Publisher and the founder and ceo of *NoahWire*
[](https://noahwire.com/)PS We’re sure you can guess how we made the image accompanying this article…
- https://petapixel.com/2025/03/26/images-in-chatgpt-ai-generator-openai/ – Corroborates OpenAI’s launch of a new image generator within ChatGPT, highlighting its advanced capabilities in producing photorealistic images and understanding of typography, color, and mood.
- https://techcrunch.com/2025/03/28/openai-peels-back-chatgpts-safeguards-around-image-creation/ – Discusses OpenAI’s updates to ChatGPT’s image generation policies, allowing images of public figures and nuanced content creation, which democratizes creativity for users.
- https://www.courts.michigan.gov/492eca/siteassets/publications/benchbooks/evidence/evidbb.pdf – Although not directly related to AI, provides insight into legal considerations around evidence authentication, which could be relevant in the context of AI-generated images being used as evidence.
- https://www.mass.gov/guide-to-evidence/article-xi-miscellaneous – Similar to the Michigan evidence guide, offers broader legal perspectives on evidence admissibility and authentication, relevant when considering AI-generated content in legal contexts.


