There’s long been a misunderstanding that a design system is a synonym for a style guide or a component library whereas these are most of the most tangible artefacts of the system.

With the natural hype, discussions and experimentation with various AI tools, we’re seeing their importance, not just as collections of things but the processes and workflows, governance and decisions and needing to make this more explicit. All of this is good not just for machines but in codifying and communicating all of the unseen or ambiguous parts of design system work for humans too.

So yes…in theory with a good structured system, your AI friend can generate all kinds of things from it and likely run the system for you, so I come back to their original premise in these new waters we’re swimming in: what’s the point of design systems and how might that evolve over time?

Done well, a design system brings disciplines across an organisation or part of around a shared understanding through clear information, methodologies and shareable things — to make the output the business ultimately needs, digital products of some kind. They enable us to improve the quality of outcomes and speed to market and often a huge undertaking for the people involved in creating, selling-in and maintaining it.

When it comes to how we use technology with this, more than ever there’s a huge number of ways that could manifest from nudges and automation of rudimentary tasks through to wholesale generation of products. Each org will have a different sense of value and risk of using any of these tools. The tools themselves change and adapt rapidly as the vendors behind them scrabble to win in their respective marketplaces with a pace of change I don’t remember seeing before.

So our design systems…what are they and what might they become? I’m not going to do the future-gazing bit and tell you everything is AI because that’s pretty lazy and assumes we’ll be doing the same thing…but with help. I think with going back to the premise of design systems they can become more and lean increasingly on what used to be unseen or ambiguous.

They’re now more than ever enablers for products. Increasingly they’re brokers between different sources of truth. In catering for AI, we’re having to encode more of what we know, which ultimately is great for people and once was static documentation that people didn’t want to update or digest takes on a new, more powerful lease of life. They’re the nexus between people and machines, intent and outcomes, so yes the age-old conversation of whether ‘design systems’ is the right term to use is fair but also an exhausting segue into semantics. The tools we use should help to amplify the people to do very human things and with design systems as the focal point, I think we can do that.