Cybernetics Versus Informatics: Understanding the Type Wars | Medium.com
tags
: [[cybernetics]] [[type theory]] [[category theory]]
source
Notes
- The two competing visions that emerged in the early days of programming came from mathematics ([[Lisp]]) and engineering (Fortran)
- systems that consume data, manipulate it in terms of semantic meaning, and produce meaningful results
- systems that satisfy requirements meaningfully and adaptively in novel and uncertain environments
- Lisp was born out of an environment that sought to understand computing as solving symbolic expressions (like you do in math)
-
Usage of Lisp declined as interest in classical AI (symbolic AI) declined, as Lisp was thought of being the language of AI
- Symbolic AI stressed a neo-[[Cartesian]] model of computing, by which sensory input was processed through “brain modules”
- As Lisp faded into the background, another mathematical concept began coming to the foreground: static type checking and “programs-as-proofs”
-
[[Clojure]] was a huge innovation in the Lisp world, bringing immutability and laziness, among other things
- Clojure seeks to maximize expressivity over formal verification, although formal verification is a goal as of 2016
- Languages like [[Haskell]] and [[Standard ML]] use their type systems to allow for formal verification
-
If a dog wanted to reach something across a barrier, in the real world it would use its accumulated experience, exploitation of the environment, and experimentation to find a solution.
- The informatics approach to solving this problem would be to model the world
- the cybernetic approach would be to create adaptive systems that are capable of adapting to such a thing