Here’s a brief apples to apples strength of AM:
A friend of mine is getting into XSI. His wife is a teacher, so he was able to purchase the advanced XSI for something like $300. He called me over to his office to show me something he was working on. He had taken a walk cycle and applied it to a character. Now the character could walk in place. After saying how ‘awesome’ that was, he then showed me that with a simple expression (an expression is a mathematical relationship) applied to the character, he could now move along a single axis at the same rate as the walk cycle so that his feet wouldn’t appear to slip. Having the character turn though… that’s a more complicated expression.
Later that week after he gushed a little more about the awesomeness of XSI, I showed him the equivalent in AM. Since you’re new (the old guys know where I’m going) I’ll tell you. Same deal. You have a character. You apply a walk cycle. You also have a path (that can turn and go any direction, left right up down). Using a handy path constraint (no expressions) you simply tell AM how long you want the character to take to get from the start to the finish. There is a similarity in that you must know how far the character is supposed to travel in one step, but instead of applying that knowledge to an expression, you just put that number in a field called (get this) stride length. Now when ever this action is applied to a character on a path (remember from before?) AM will know to move the character that amount down the path, whichever way it goes.
AM’s biggest strength, as I’ve come to see it, is that it takes very high end concepts (read: those also found in more expensive software) and makes them usable. Under the hood, it’s got to be every bit as complex as any other piece of software out there. However, what AM shows the user through the interface is usually a well thought out, elegant solution that strips out the complexity that remains flexible.