What happens when anyone can generate a ‘good’ design in seconds?
For decades, design has been defined by craft – mastery of tools like Photoshop, Illustrator and Figma and an understanding of the fundamental design principles, including layout, typography and visual storytelling. But the rapid emergence of AI tools such as GPT, Midjourney, Runway, Claude and Adobe Firefly is shifting that foundation.
Design is no longer just about making things. Increasingly, it’s about directing machines to make them.
This shift is not only changing how designers work – it’s redefining what it means to be a designer in the first place.
When we started these workshops in April 2024 - just one year after ChatGPT had been released - confidence among the attendees was actually higher than usage frequency. People felt more capable
than active (and the rollout of Copilot or other tools to corporate teams was slow, so there was probably
an access issue too).
Traditionally, designers were hands-on creators. Every pixel, layout and asset was crafted manually. Today, AI tools are transforming that role into something closer to a creative director.
Some designers now put the sketching stage as a lower priority in the design process, instead opting for the prompt box for the initial brainstorming. Mood boards previously involved scouring the depths of tools such as Pinterest, Google and Dribbble to find design inspiration. Now boards can be generated instantly using tools like Midjourney. Video concepts would take hours, sometimes days, to put together, manually animating or storyboarding to help convey a concept. Now platforms like Runway allow for the easy generation of high-quality video to give pitches a more polished feel. UX has also seen a shift – there was a time when a UX copywriter was a full-time role; now this is becoming increasingly obsolete with tools such as GPT or Claude generating highly effective UX copy and structures.
The skill is no longer just execution – it’s curation, judgement and direction.
This raises important questions: Is prompting becoming as fundamental a skill as learning a dedicated design tool such as Figma or the Adobe Creative Suite once was? And is prompting the only skill a designer needs for the future?
One of the most immediate impacts of AI is speed – particularly in the ideation phase.
What used to take days – brainstorming, sketching, exploring variations – can now be done in minutes. Designers can generate dozens of visual directions almost instantly, iterate rapidly, and present a wider range of ideas to clients or stakeholders.
In practice, this is already changing workflows:
But this acceleration introduces tension. While speed can unlock creativity, it can also dilute it. AI-generated outputs often fall into what has become recognisable as an ‘obvious AI’ aesthetic – particularly in the rendering of human faces. As a result, many visuals begin to share similar characteristics, making them feel less unique and more generic. When ideas are abundant and easy to produce, the challenge shifts from creating options to choosing the right one. As AI tools enable designers to generate a far greater volume of creative options, clients are increasingly faced with an overwhelming number of choices. This reflects the ‘paradox of choice,’ a concept introduced by psychologist Barry Schwartz, which suggests that having more options can make decision-making harder, not easier. Instead of increasing satisfaction, too many choices can lead to decision fatigue, slower approvals, and greater uncertainty about whether the ‘right’ option has been selected.
Perhaps the most controversial aspect of AI in design is whether it’s weakening foundational skills.
There’s a growing argument that AI tools may encourage surface-level design. Designers may begin to rely solely on generated outputs rather than building things from scratch using their own creative skills. Where these tools are not available in certain cases, will designers struggle to think independently? The fundamental skills of typography, composition and colour theory will become something confined to old design textbooks, with no perceived use for learning these skills. There is also less incentive to master complex tools such as Figma or the Adobe Creative Suite when AI can produce something in a fraction of the time, without the need for research and practice.
The risk is the emergence of designers who can generate visually appealing work, but who lack the underlying knowledge to refine, troubleshoot or innovate beyond the tool’s output.
On the other hand, AI can be seen as removing repetitive, low-value tasks:
This frees designers up to focus on higher-level thinking:
This leads to a deeper question:
Is knowing how to design becoming less important than knowing what good design looks like?
So how does all of this affect the real-world job market for designers looking for new roles?
Design roles are evolving, with employers beginning to expect familiarity with multiple AI tools and workflows. The ability to integrate the use of AI into the creative process is essential to be comfortable with the expectation of rapid iteration and experimentation of concepts.
At the same time, teams are becoming leaner. A single designer equipped with AI tools can now accomplish what previously required multiple specialists.
We are also seeing the emergence of hybrid roles:
In some cases, companies are even bypassing traditional roles altogether – using tools like Midjourney for branding exploration or AI-generated assets for marketing content instead of outsourcing. This sparks a lot of internal company conversations as to the necessity for the designer role going forward, making the industry feel understandably nervous.
AI has dramatically lowered the barrier to entry in design.
The upside
Non-designers can now create high-quality visuals, which leads to more people feeling empowered to be involved with creative work, rather than leaving it completely to the role of the designer. This then means smaller teams are still able to produce professional outputs without the need for large budgets.
The downside
The market is becoming saturated with ‘good enough’ design. It’s harder for designers who have worked hard to craft their skills to stand out. The differentiation between authentic and generated design is less clear than ever. Thus, the value of basic design execution is rapidly declining.
This creates a fundamental tension: If everyone can design, what makes a designer valuable?
The answer increasingly lies beyond tools – in taste, judgement, originality, and the ability to solve real problems.
Rather than replacing traditional tools, AI is being integrated into broader, hybrid workflows.
A modern design process might look like:
Design is no longer about mastering a single tool – it’s about orchestrating multiple tools effectively.
This shift places greater emphasis on adaptability and experimentation. The most adaptable designers are not those who rely on one platform, but those who can navigate an evolving ecosystem of tools.
AI is developing at an extraordinary pace, and its impact on the creative industry is undeniable. But the narrative that it will simply replace designers is overly simplistic.
What’s happening is more nuanced.
AI is redefining the role. Increasing the efficiency of repetitive tasks and placing more importance on the strategic aspects of design, the direction and decision-making. Expanding the expectations of what multi-disciplinary designers can achieve.
For designers, staying up to date with emerging tools and experimenting with different workflows is crucial to stay relevant in a constantly shifting market. This should be coupled with the continuation of foundational skill development that underpins good design.
The designers who will be the most resilient to inevitable change are not those who ignore AI, nor those who rely on it entirely. They will be the ones who integrate it thoughtfully, using it to enhance efficiency and creativity, while maintaining the skills and perspective that give their work meaning.
AI is not the end of design. It’s the beginning of a new version of it.
I will leave you with this to ponder: if this article was written with the help of AI, does that diminish its value – or does it prove the very point, that efficiency is becoming just as important as authorship?