I read the first 100 pages or so of Philisophical Investigations before giving up since I had lost track of the lines of reasoning. Here's what I got out of what I did read though:

Language is far more important in philosophy and life than we think. The biggest takeaway I got from the part of this book I read is a way of looking at problems with the limitations of language in mind. Ask yourself: to what extent is thinking speaking? What is thinking without language? Is it even possible, or is that something else?

There's this interesting idea that has stuck with me, which is that when we say words, "pictures" are "brought before" our mind. Is language (speaking, reading/writing) simply a way for us to conjure up these mental images in other people's minds? If so, how can we be sure that what they see is what we intended for them to see? I think it's clear from experience that the images are mostly right, most of the time. But when they're not, we have misunderstandings. Another interesting statement made in the book (iirc) is that we only need more language when we feel there is a misunderstanding. The word "more" is important in the previous sentence. The idea here is that when we speak, we have some desired outcome from the outset. Once we feel that our speaking has led to the outcome we wanted, we are satisfied to stop speaking. It is only when the person we're speaking to isn't doing what we want, or seems to be getting the wrong mental picture that we need to continue speaking (this is what I meant by "more language": to continue speaking).

There's another interesting area Wittgenstein explores (which I can't admit to following very well), but I'll try to conjure a mental image in your head of it ;). Basically, so far his argument (if I understood correctly) is that "the truth" is the mental images we see and the actions we take, and language is just a means to those ends. He then argues (again, if I understood him correctly) that we usually run into trouble when we take language as the starting point of knowledge. That's not very clear, so what do I mean by that? It's sort of like, words work well when we're using them to achieve some outcome. But they start to confuse us when we use them without a desired outcome from the outset: when we use them to gain knowledge. Words are not facts that we can logically make deductions from to discover new knowledge.