You may have encountered the term 'generative design' in the context of producing design permutations, creating geometry from some simple inputs, or even just building computational graphs using Dynamo or Grasshopper.
We see generative design (/ gen·er·a·tive de·sign / noun) as:
A collaborative design process between humans and computers. During this process, the designer defines the design parameters and the computer produces design studies (alternatives), evaluates them against quantifiable goals set by the designer, improves the studies by using results from previous ones and feedback from the designer, and ranks the results based on how well they achieve the designer’s original goals.
Above: Some generated alternatives - Mars Innovation District - The Living
Generative design is a specific application of the computational design approach, with the following distinctions:
The designer defines goals to achieve a design (rather than the exact steps).
The computer helps the designer to explore the design space and generate multiple design options (not just one).
The computer enables the designer to find a set of optimal solutions that satisfy multiple competing goals.
The designer compares multiple design scenarios to find a set of design options that fits the design goals.