The simplest computer
A Turing machine is a head on an infinite tape of cells. At each step it reads the symbol under the head, and depending on its current state and that symbol, it writes a symbol, moves one cell left or right, and switches state. That is all, and yet it can compute anything any computer can: the Church–Turing thesis says this model captures exactly what "computable" means. Writing small machines by hand is how theory-of-computation courses make that concrete.
Reading the rules
The rule q1, b -> q2, Y, L means: in state q1 reading b, write Y, move left, and go to q2. The machine halts when it enters an accept or reject state, or when no rule matches the current state and symbol, which counts as rejecting. The tape starts with the input written from the head rightwards; every other cell is blank.
The examples
aⁿbⁿ shows the standard marking technique: replace the first a with X, run right to the first b and replace it with Y, run back, and repeat. If every a finds a b and nothing is left over, accept. Binary +1 walks to the end of the number and carries leftwards, turning trailing 1s into 0s until it can write a 1. Palindrome erases the first symbol, runs to the end to check the last one matches, erases that too, and repeats from both ends inwards.
Halting and looping
Unlike finite automata, a Turing machine might never stop. The simulator runs up to 5,000 steps and reports "did not halt" beyond that. No program can decide in general whether a machine halts (the halting problem), so a step limit is the honest answer. Machines here are deterministic: one rule at most for each state and symbol.
Where this fits
Turing machines sit at the top of the hierarchy. Below them, the PDA simulator covers context-free languages with a single stack, and the NFA & DFA simulator covers regular languages with no memory beyond the current state. Comparing what each machine can do with the same language is the quickest way to see why the hierarchy has the levels it has.