You Are Confusing the Program with the Data.

Being within the kind means that genetic information was not gained. In other words, you cannot, through chance, cross the line between change within the allotted kind and the addition of more information that would be required to form a specified change beyond the original design capabilities (racehorse breeding for example).

To boil this down to layman's terms, the original kinds have the ability to diversify already designed into them, but this diversification takes place only within the originally designed limits ("kind").

To use a model, it's like a computer program and a computer program's working data.

You can configure a program to change the way it works within it's predesigned limitations by changing the (genetic) data you provide to it, bur you absolute cannot set a computer in a office for 6 billion years and expect the program itself to change in any advantageous way. It will only statistically deteriorate.

It is easy to confuse these two aspects of diversification and so it (this easily induced confusion) is used as a false premise to induce people into believing a particular change is actually an evolutionary process when it is actually the expression of a predesigned parameter.

As ironic as it is, the additional complexity required by a self-adapting system (such as we now clearly see and agree upon) dramatically increases initial design complexity and therefore decreases the already astronomically low (impossible really) chances of evolutionary processes being a viable explanation.

Hopefully this makes sense to everyone.