A good design system shouldn't maximize choices. It should minimize unnecessary decisions.
In his famous TED Talk and book "More is less", The Paradox of Choice, Barry Schwartz challenges a common belief:
More choices → More freedom → More happiness.
But having more options comes with unexpected side effects:
- Decision paralysis: Too many alternatives make it harder to choose.
- Regret: Every choice reminds us of the opportunities we left behind. What if?
- Higher expectations: More possibilities make us less satisfied, even when we've made a great decision.
So, what does this have to do with design systems?
Quite a lot, actually.
We often confuse flexibility with quality. More component variants, more properties, more configuration, more customization. But a component with 50 configurable properties isn't necessarily better than one with 5.
Every additional option introduces another decision, another potential inconsistency, another opportunity to get something wrong.
And this matters even more in the age of AI where it can generate endless interface variations in seconds. But without shared principles and constraints, more possibilities simply mean more unpredictable outcomes.
This is one of the core ideas behind ViraUI (https://viraui.dev), the AI-centric design framework I'm building, finding the perfect balance between:
- Flexibility to create unique experiences.
- Predictability to make outcomes reliable.
- Consistency to scale across products and teams.
By giving designers, developers, and AI agents the right constraints, we can reduce decision fatigue without sacrificing creativity, while keeping high standards.
Because constraints aren't always limitations. Sometimes, they're what makes freedom possible.
Top comments (1)
The Schwartz connection lands well here. The point about AI generating endless interface variations makes the decision paralysis argument feel current rather than theoretical, since the cost of adding another variant is now close to zero. I also liked the three-way balance for ViraUI: flexibility, predictability, and consistency. Predictability is the one that usually gets dropped when a system chases flexibility, so it is good to see it named as an equal. Do you find that the predictability part is what teams push back on most when they first adopt the constraints?