Evolution is adaptation, it doesn't make sense to say that there is adaptatation (if you are going to admit that it is due to genetic change) but not evolution.
Randomness can lead to things that are more perfect than human beings, at least at certain jobs. See pretty much the whole of the big parts of the AI industry - None of the best AI are designed by hand. The only thing designed by hand is the environment the AI are tested in, as well as a system to generate random AI:s and a system to "mutate" AI:s. Things that affect evolution, but that are not evolution themselves. So you can't use these select few intelligently designed parts of the development of AI as an argument that evolution requires intelligent design, the best argument you can make is that things had to start out with intelligent design. While AI:s have a long way to go before being as versatile as humans are, a lot of the best AI today completely outmatches humans at at least one task. So the argument that randomness can't lead to something very functional is wrong.
If one organisms dna changes, the organism as a whole will be slightly different from the other of it's species. That is not silly, is it?
Now, one change of something times a few thousands is a total change that is greater than the lone change. This isn't silly either, right?
It's basic math. Say I have two dots on the same spot on a coordinate system, and I continually update the dots by adding a really small random vector. The dots will have really similar positions in the beginning, being quite indistinguishable. But over time there is a possibility that the dots will part ways and eventually their positions are going to be really different. With every new iteration, the "old" dots and the "new" dots are basically the same, but the difference between the origin dot and the ten millionth dot could be really big. A lot of small changes together make a big change. That is evolution. Each change is not enough to cause speciation, each change could be what bhs and d1 wants to call nothing but an "adaptation", but what stops the new adapted organism from going another step in the same direction to better adapt? And another one? It's basically the same species with every new step, but at some point the accumulation of changes in one direction will be quite large. Unless you want to argue that the adaptations are infinitesimals, it's impossible for repeated adaptation to not lead to larger changes in total. So it's plain ignorant to call the idea that some huge number times a small number can equal a big number "silly" (a lot of small changes will lead to a substantial change).