Would alien or artificial life do things very different from us? Does our human form limit how we do things?
(Based on material from TheoreticalRigorCantReplaceEmpiricalRigor)
Be careful there now. Computers are not replacements for humans, and in many people's opinion should not be replacements for humans. The models that humans should use for computing should be for computing and processing batch tasks and data, not for building humans or models of humans. Those in the Lisp/AI field (isn't that dead now?) can argue all they want - but we have plenty of humans in China and USA (and other countries), so we don't need no effing more selfish pricks on this planet trying to take over the world. What we need is automation, and free slavery, according to AntiCreation and SlaveOrientedProgramming.
There seems to be some context confusion. If we want to test the value of a technique/tool on *human* programmers, then matching humans does matter. I perfectly agree that future AI may prefer BrainFsck or something else equally goofy (to humans). But the assumption is that we are testing stuff for human developers. I'd expect that a being that lived in a 4D universe (real or emulated) may approach programming and databases very different from us. They may have 3D printouts and editors for their code. (Although such is technically possible for us, it is not practical because things tend to overlap and get in the way of each other.)
We might use 2D printouts, but both our languages and our computers are generally very one dimensional. The dimensions we use to make a tool are generally the minimum needed for the problem space: one dimensional memory because it makes indexing easier, and higher dimensional memory can be implemented via offsets; one dimensional code because time is linear, and most nonlinear time models can be emulated in a linear model anyways; two dimensional tables because we are listing identical objects, and then grabbing attributes of an instance, which is essentially just a binary operation, and hence requires a two dimensional lookup table. It is not like we don't delve into more complex paradigms like trees and graphs and atomic operations and whatever when needed.
Yes, but 3 or more dimensions are not natural for us to work with in terms of visual I/O. Even in CADD, most of the actual design work is done in 2D or limited 3D. A lot of our IT structures form graphs, but we usually project them into 2D because labels among other things tend to overlap in a 3D model.
Would their Dr. Coddd (3 D's = pun) use 3D tables?
If a deep-sea creature that only used sonar to navigate instead of sight grew intelligent, what kind of "screen" would it use?
would it even have the concept of a 'screen'?
Some kind of device to project or emulate models and symbols.