Last week I talked about LLMs as "spackle for toil".

The original software-based spackle for toil is spreadsheets.

Spreadsheets are absurdly, generically useful, in just about every domain.

They are many orders of magnitude better than doing the calculations by hand or on paper.

This immediately changed a huge class of problems from "infeasible" to "possible".

Spreadsheets are a platform for calculation; they are inherently messy because they are so general purpose.

They can absorb any given domain in a slapdash, jury-rigged way.

Once a given use case becomes particularly common, it makes sense to lift the use case out of a spreadsheet and into a bespoke tool with use-case specific rails.

But you can think of spreadsheets as the messy / good-enough fabric for a lot of use cases to have their MVP.

Spreadsheets set the level of motivation necessary to build the MVP much lower, enabling many more good ideas to be tried.

Spreadsheets will never go away; they will remain forever in that part of the funnel of creation.