The fundamental idea that Kant introduces in the second half of Chapter 2 is that rational beings are "ends in themselves." When you settle on a course of action, Kant notes, you do not think of yourself as a means to some other purpose; you think of yourself as the purpose or "end" to which all your actions are directed. If you expect other people to accept your motives, you must respect the fact that other people also think of themselves as more than mere means to other goals. Thus your motives will lack universal validity unless you respect the fact that all rational beings have intrinsic worth, just as you do. The categorical imperative requires you to treat all your fellows as "ends in themselves"--that is, as objects of intrinsic value--and not as mere instruments for the attainment of your personal goals.
Kant's four examples of duty are no more successful in substantiating this idea than they were with the categorical imperative in the first half of the chapter. (Does failing to cultivate our talents really violate our notion that all people have intrinsic worth?) Nonetheless, his core insight fits fairly well with most people's basic sense of morality. In practice, Kant's notion of the "moral law" and the categorical imperative sounds a lot like the Biblical doctrine that we should treat other people as we would like them to treat us. Similarly, his notion of people as "ends in themselves" fits with the modern idea that all people possess a fundamental dignity. It is wrong to abuse people, or enslave them, or use them for selfish purposes, because doing so violates our sense that people are not physical objects that we may use as we see fit.
Kant's notion of a "kingdom of ends" also fits fairly well with modern ideas about politics. Though Kant is writing about morality, not politics, his description of the ideal community as one in which all people create their own laws is in essence a picture of democratic society. In practice, of course, societies must make laws by balancing different interests and viewpoints within a constitutional framework. In theory, however, democracies are based on Kant's notion that laws are valid if and only if they make sense to the people who must follow them.
Nevertheless, Kant's position is again vulnerable to the criticism that it is too abstract to be useful. Kant seems to think that reason is something static that people can use to develop universal laws and principles. In fact, different ideas make sense to people at different historical times and in different cultures. Kant seems to think that the notion that people are ends in themselves can provide clear moral guidance. In fact, this principle could be used in support of different viewpoints. (To pick just one controversial example, does abortion treat a potential baby as a mere means? Or would banning abortion treat women as mere means to the creation of babies?)
Kant's notion of "autonomy" is similarly suspect. Admittedly, Kant concedes that his notions of "autonomy" and of a "kingdom of ends" are ideal concepts that we cannot expect to encounter in real life. Still, we may want to ask whether it makes sense even to try to imagine a person making decisions without reference to any personal experience, cultural assumption, or material interest.