George Rebane
[Reading some recent papers on the current problems in the advance of deep learning, and causal inference as the next quantum leap in machine intelligence, I ran across a lot of discussion about what are still called algorithms. It recalled the summary of a more facile syntax that I and several others have been promoting. More to be said later about the contention between the ‘data fitting’ of neural nets, and ‘data interpretation’ capability of causal nets.]
al·go·rithm n. A step-by-step problem-solving procedure, especially an established process for solving a problem or accomplishing a task in a finite number of steps. (more here)
When we relieve ‘algorithm’ of its codal baggage, we are left with the simple, compact, communicable, and elegant ‘algo’. In this form the entire enterprise and field of algorithmics (better ‘algoristics’) can be efficiently communicated, bypassing current attempts to twist its ‘rithm’ tail into tortured forms like “algorithmatized” or use many other words to describe what can be expressed directly. Consider the following little dissertation on the subject.
Any process is algorizable and can be algorized by an experienced algorist. Algorized processes benefit from being readily teachable and passed on to others. Developing algos is also an art because there is usually more than one algo that can algoristically capture the essence of a process. Algoristically speaking, most algorized processes can always benefit from additional algorizing. The history of algoristics is full of such algoristical efforts of developing ever more efficient algos. In the affairs of Man never disparage algorization, for it simply is the heart of the matter. No light bursts forth in your mind save in an algorizable form, for without the ability to algorize we proceed as the blind. It always pays to be an algophile instead of an algophobe, for expressing and retaining knowledge, it all comes down to algoristics.
One of history’s greatest algorists was Henry Ford. He took the notion of interchangeable parts (first developed for the manufacture of military arms) and algorized the manufacturing of cars into the assembly line. Realizing the algorizability of automobile manufacturing, his algorized assembly line was algoristically so structured that parts didn’t come to the assembled car, the car went to the parts. That was the breakthrough algo that changed a country and then the world.
- Algo
- Algorist
- Algoristics,
- Algoristical
- Algoristically
- Algorize
- Algorized
- Algorizable
- Algorizing
- Algorization
- Algorizability
- Algophile
- Algophobe
Comments