@deodad

the output of design used to be visual artifacts many attempts were made to turn this into code (e.g. no code, wysiwig) but the gap between visual artifact <> code was never satisfactory closed because there are too many unspecified variables LLMs have closed the gap—designs are building specifications (e.g. serialized figma nodes) and LLMs implement and adapt into whatever environment / language a consequence is there is more responsibility on designs to organize the internals like tokens and components and positioning (i.e. the specification) giving rise to the visual artifact because that's the layer the LLMs are using to implement if they do this well the FE engineer's job becomes almost focused on backend for frontend

1 reply

1 recast

10 reactions