>>13118621Turing answered the question of "which problems are computable problems" by defining a machine which moved along an infinitely long ticker-tape, and it reads what's written on the tape and has some rules for what to do when it reads certain things on the tape. The machine is called a Turing machine and a computable problem is "what a Turing machine computes" and "turing complete" means capable of doing everything a Turing machine can do, that is to say, capable of simulating anything in the universe (with enough memory and time), including your laptop and your brain.
The issue is that Turing machines are obsolete, one should use RAM machines, and the best model is just the machine you are reading this on, idealized to infinite RAM. That's it, that's a RAM machine, and it's equivalent to a Turing machine.
The original Turing machine itself is kind of nice, because it is extremely simple, but any basic computer is equivalent to any other, so you only need to build a path from A to B by having A simulate B and B simulate A. That's all. Today, there are simpler known universal machines: Von-Neumann's autamata, Conway's, Wolfram and Cook's. These are the simplest to date. But pick automata at random, a certain fraction will be universal, that's something appreciated by all three.
The tape bullshit is just standard mathematical obfuscation. There is no gain in precision or clarity from the formalism.
