In July of 2019, we released a modest start to Components AI. We've been working on exploring two questions:
If we lived in a world where every possible iteration of a design was already built - what would your workflow be to define what you’d ship? What tools would you use to sift through all of this generated noise? A common critique around generative / procedural design output is that it’s too overwhelming. Too many choices. This is a valid consideration when thinking about the ergonomics of a tool or interface. But, through a different lens, the overwhelming amount of potential solutions exist, regardless if a piece of software is generating them or not. What do we need to make total generative space more manageable to explore?
We think the future of design tooling will be centered around a search interface. One that allows you to filter or define the constraints and quickly cycle through available outputs.
We want to build a design tooling platform that allows you to generate, assess, and test designs at unprecedented speed and scale. Whether it's visualizing a wide variety of content choices, theming compositions, or layout options, it should be effortless to generate and explore design variants with your team.
We believe that successful design in the future will necessitate creating things beyond our individual comprehension. This will require us to create more resilient and dynamic systems that learn and respond to feedback over time.
We currently have a lot of questions but we’re excited and energized about the opportunity to explore potential answers.