An evolving story of innovation

Artificial intelligence is changing every industry rapidly, but this shift is evolving in a way we have not witnessed before. Designers have key skills worth noting: problem solving, agility, curiosity and, most importantly, creativity. In a future world of robots and AI tools surrounding daily routines, creative people should be able to adapt. Human intelligence is conceptual and vision-led: designers begin with an idea in mind and develop it through judgement, while LLMs predict the next step in a process without that originating vision. We will continue to orchestrate - the human in the loop. Some design will be automated, but designers will still instruct these processes, critique the output and improve the results. Designers will be needed to work alongside AI.

A good place to start is by widening your perspective to the possibilities ahead. After working as a designer across websites, apps, campaigns, branding and furniture for 25 years in London, I have built up a lot of experience, but I was often reliant on others to build or take my ideas to the next stage. That has changed. We can now realise what our appetite feels like. The skills start with prompting: the more you experiment, the better you can control the outcome. Continue to design for humans, but also turn your thinking upside down and imagine what AI needs to see to understand your vision. Think about how things are drawn using maths. Play with prompting SVG graphics or tools like Figma Make to generate ideas. AI creates by predicting what happens next, so you need to define frameworks, criteria and constraints, including what not to do. Once you refine your ideas through prompting, embed what you have learned through summaries, skills and memories, as you would with AI agents.

The production side or repetitive bits can already be done - albeit with minor tweaks from the designer, which is good news. Think of Photoshop retouching work: removing shadows, blemishes or unwanted details can now be done in seconds. Image generation has already affected advertising, and AI-supported research and early-stage client ideation are becoming commonplace. UI design and digital experience will become much more dynamically created. Imagine webpages generated on the fly to serve the content you need from a prompt. Chat surfaces will blend with webpages, allowing people to use multiple modes of communication to retrieve information and learn what they need.

Maybe this is not the most relevant question, because we need to rethink what it means to be creative. Our design work is shaped by our culture and environment. As Yann LeCun said in a recent interview, “Reality is way more complicated than language. It is noisy, and training AI to understand reality is really hard.” So think about that: can a prediction machine really imagine like you can? LLMs do not plan a task or determine an outcome. If your work is highly repetitive and lacks rigour, it may be automated before the end of this year. Someone who is a little entrepreneurial will flourish, possibly reducing the need for a job in the traditional sense. If you have ideas and taste, you will continue to create in this world, whether you have a job or not.

Source: Yann LeCun interview on AI and reality

Here lies the challenge, many design ideas come from concepts or visual references deeply planted in our minds. Experiment by using words to describe your intention. Digital design or other creative fields may be easier to control, however the models are improving. Use voice and text to try prompts and switch between models to compare outcomes. Start by listing the qualities of your design idea in simple language. How would you describe your work? A markdown file explaining the design language for a website or app can also lead to surprising results when you ask an LLM to build.

The soul of design: the quality that makes an idea resonate or shine. This is the biological skill humans possess, shaped by evolutionary pressure, environmental change, social behaviour and an advanced brain. We know how to design for humans. Now we need to pivot toward allowing machines to support our imagination, while keeping human judgement at the centre.