… is that he doesn’t suggest actions, he just rules them out..
His Categorical Imperative is a test for motives — roughly. The idea being if the action you propose to take (really, the maxim upon which you are going to act) passes the Categorical Imperative, it’s morally permissible.
The trouble is, that it doesn’t suggest what you should do — only what you can’t do. So, if you have no maxims for action that pass, what are you supposed to do? Especially if non-action is one of the possible maxims that fails. So, Kant doesn’t help make decisions in really bad situations.
Contrast this to Utilitarianism, which gives you a way to measure ALL proposed actions — even if all of your choices increase pain, you can choose the one that increases the pain the least.
Another problem with the CI as essentially a Veto is choosing between several maxims that pass — then what should you do? Kant’s response is pretty much, figure it out yourself –do whatever seems better in some other way.
Contrast this with Virtue Ethics, which is supposed to give a sort of holistic way to be a good person. You can always choose an action that increases virtue — and you have a set of characteristics of a good person to go by in your general life.
All that being said, Kant is still one of my favorites. Really, I like the folks who followed (and translated) Kant — they improved on his idea in attractive ways.