Nice job including Excel! The world's most used programming tool! Having worked on that from as early as 1992 and as recent as 2018 I can tell you it is loved and used for everything from poetry to large-scale financial applications on Wall Street. Another useful tool used for long tail office productivity solutions was Visual Basic for Applications (VBA). Each Office application created an object model that exposed their capabilities inside Visual Basic. A myriad of 'emergent' solutions could be built satisfying specialty needs to the right of the tail built on more generalized capabilities found on the left of the tail. VBA hit some security snags in the early 2000s and VB as a language died out. But elements of this idea are re-emerging in the cloud version of Office using JS.
Love this example! There's a whole class of programming languages that are so similar to English that they begin to meet criteria #3 here. SQL is another that comes to mind. Thanks for sharing!
This is a great read, thanks for the clear articulation of this fairly nuanced topic!
My 2 cents at the risk of stating the obvious: designing for emergence leads to more expensive implementations than user-centered design. These meta designs that allow users the flexibility to design their own experience require building user controls for things that could otherwise be "hard-coded".
So the long tail can feel being stuck between a rock and hard place - (a) making incremental improvements that may not justify the cost and (b) taking on the risk of building something expensive and flexible, hoping that users find their own ways to use it. Do you have more insights to offer here? :)
This is a great question! My answer may not be satisfying but it's the best thing I've seen work — build the 'blocks' as scrappy as possible, make it accessible to highly motivated users, and then pay close attention to when/where/how they use it in order to help guide further refinement.
There’s also a third state in between end users controlling things and the designer controlling, which is to have a subset of users who are technical build stuff for the other end users. Basically to have the market build for you (but it’s not only done by people looking for financial gain). this is a very popular and successful technique with stuff like iPhone’s App Store (or computer software in general), Minecraft/roblox/vrchat servers/worlds, browser extensions and browser websites. Notion has some users selling their templates for others to use. Obsidian notes has a collection of community made extensions as well
Great post and it sparked a lot of interesting questions for me.
In particular, how do we know we are on the right path to building an emergent set of "things?" It feels like the new properties of the emergent system would be detected as "surprises" to the original creators. Maybe there is some ratio that you need to get on the other side of to say that there are strong emergent properties? But this would be a following indicator rather than a leading one...
Nice job including Excel! The world's most used programming tool! Having worked on that from as early as 1992 and as recent as 2018 I can tell you it is loved and used for everything from poetry to large-scale financial applications on Wall Street. Another useful tool used for long tail office productivity solutions was Visual Basic for Applications (VBA). Each Office application created an object model that exposed their capabilities inside Visual Basic. A myriad of 'emergent' solutions could be built satisfying specialty needs to the right of the tail built on more generalized capabilities found on the left of the tail. VBA hit some security snags in the early 2000s and VB as a language died out. But elements of this idea are re-emerging in the cloud version of Office using JS.
Love this example! There's a whole class of programming languages that are so similar to English that they begin to meet criteria #3 here. SQL is another that comes to mind. Thanks for sharing!
This is a great read, thanks for the clear articulation of this fairly nuanced topic!
My 2 cents at the risk of stating the obvious: designing for emergence leads to more expensive implementations than user-centered design. These meta designs that allow users the flexibility to design their own experience require building user controls for things that could otherwise be "hard-coded".
So the long tail can feel being stuck between a rock and hard place - (a) making incremental improvements that may not justify the cost and (b) taking on the risk of building something expensive and flexible, hoping that users find their own ways to use it. Do you have more insights to offer here? :)
This is a great question! My answer may not be satisfying but it's the best thing I've seen work — build the 'blocks' as scrappy as possible, make it accessible to highly motivated users, and then pay close attention to when/where/how they use it in order to help guide further refinement.
There’s also a third state in between end users controlling things and the designer controlling, which is to have a subset of users who are technical build stuff for the other end users. Basically to have the market build for you (but it’s not only done by people looking for financial gain). this is a very popular and successful technique with stuff like iPhone’s App Store (or computer software in general), Minecraft/roblox/vrchat servers/worlds, browser extensions and browser websites. Notion has some users selling their templates for others to use. Obsidian notes has a collection of community made extensions as well
Great post and it sparked a lot of interesting questions for me.
In particular, how do we know we are on the right path to building an emergent set of "things?" It feels like the new properties of the emergent system would be detected as "surprises" to the original creators. Maybe there is some ratio that you need to get on the other side of to say that there are strong emergent properties? But this would be a following indicator rather than a leading one...
Anyways, really interesting!