Collaboration is either generative or reactive

Collaboration comes in two forms, distinguished by when it happens relative to the act of creation:

• Generative collaboration happens when we’re all part of the initial act of creation: writing in tandem, pair-programming, sketching on a whiteboard, building a schedule for the offsite in real time.
• Reactive collaboration happens after initial creation: the creator solicits feedback on her creation in order to improve or validate it.

Both are valuable, but they’re not interchangeable:

• Generative collaboration produces shared work and is generally better suited for change management, as people feel stronger ownership over something they helped create.
• Reactive collaboration refines work that already has an author.

The choice between them shapes both the quality of the outcome and how much the people involved feel ownership over it.

There’s reason to think AI is shifting us from generative toward reactive collaboration. As the cost of production collapses, it becomes cheaper for one person (plus a model) to produce a complete artifact and then solicit feedback than to co-create it from scratch. We increasingly generate first and validate downstream—which is a move from generative to reactive collaboration. That’s efficient, but it may also forfeit some of the shared ownership that only generative collaboration produces.