The basic idea of philosophy is to examine those things that are normally taken for granted. This should make it the most critical approach — critical meaning just that, to really consider what the right answer is rather than just accepting something without looking into it, as they say, uncritically. Yet, those using empirical sciences to answer some question can often say with reason that the question should be solved empirically, not by mere philosophical speculation. Suddenly, it is philosophy that is too naïve and not critical enough. How can it be so?
Speaking very broadly, philosophy has no method other than reflection, thinking about things, largely trying to clarify what is already in your mind. You can’t find out things this way if you don’t already have the necessary information about what you’re trying to figure out. In empirical questions, you have to apply the scientific method of going out and looking at how things actually are. Thus, when we have the possibility of empirically checking how something really is, just thinking about it is a completely inferior way of trying to find out the truth. Philosophers also sometimes misunderstand the limits of what they can reasonably speculate about and think that they have more or less proven ideas that are completely unprovable, like other universes where alternative possibilities exist, or a self-particle in the brain to explain the continuity of the self. The world is what it is, and most of it is not in your head already. You can’t derive it all by logic, and you’re liable to just rationalise your biases if you try that sort of thing.
In such cases, the philosophers are being naïve and providing entirely the wrong kind of answers. However, sometimes it is the other way around. Empirical scientists can be naïve for lack of philosophy, especially when trying to answer philosophical questions. Philosophy isn’t just a field, it’s an overall approach to anything and everything, and those who really understand their field also understand its philosophy by definition. The philosophy is the underlying principles that they understand. The philosophical method here is to look at and see the those principles, and also to distinguish what you’re talking about at a given moment.
Thus, suppose a natural scientist sets out to empirically establish whether there is free will, and concludes that there isn’t because human actions are deterministic. Then he thinks the whole question has been answered. Actually he just missed the point. He though he already knew what free will was, but if he understood the philosophical side of it, not to mention if he had studied the philosophical literature, he could have seen the big question (or the controversial question at least) is what free will is anyway. Philosophers arguing about free will don’t spend most of their time arguing (unempirically) whether human actions are deterministic or not. They talk about whether that would mean there can or cannot be free will. Indeterminism is fundamentally random, after all — while we seem to want to avoid determinism to be free, do we want that either? There are also other perspectives on what constitutes free will, unrelated to determinism. A more realistic empirical examination might conclude that our freedom is compromised because we are irrational and often controlled by our unconscious. And even after that, there’d be something further to consider.
Another example of a naïve scientist could be one who claimed that the problem of what consciousness can be easily answered by explaining how the brain processes information. The so-called hard problem of consciousness isn’t even about that, though, but about why there is such a thing as subjective consciousness at all — why isn’t it all matter, causality, and/or information, why is there the subjective feel as well? You don’t answer this question if you don’t consider it at all.
Perhaps the worst case would be someone who claimed that since she’d shown some behaviour to be natural in an evolutionary sense, it was morally acceptable. If you ask why this would be bad, consider that just about anything could be “natural” in such a sense, such as genocide. Ethics, a branch of philosophy though also a practical art, would be needed for saying what we should actually consider right. (See also this.)
Even people who seem to really understand their own field in science seem to fall trap to such blunders when it comes to difficult questions dealt with in philosophy. This speaks for the usefulness of philosophy as a field as well.
It’s one partly true stereotype about philosophy that all it can do is pose questions, not answer them. Even if that were so, you need to know what the questions being asked are in the first place if you want to get anywhere. Though we must remember it can’t replace empirical studies, philosophy is needed — because it’s the art of figuring out what the heck you’re talking about in the first place.