The subtitle of this book tells you pretty much everything you need to know about its contents. Okrent, a linguist, explores both contemporary and historical attempts to create artificial languages.
For each primary language that she explores, she provides not only a brief biography of its inventor (longer in some cases, such as that of the troubled man who invented the symbolic language Blissymbolics), but examples of the language in use. So you get various phrases written in English and translated into Esperanto, Klingon, and so forth. Moreover, you often get an intermediate translation, which shows what is literally being said before moving on to the smoother, "natural English" version.
The fundamental flaw faced by nearly everyone who has tried or is trying to create an artificial language from scratch is that they want to make a language that is superior in some fashion to languages that have developed organically over time.
They want their new languages to be more logical, more emotional, simpler to learn, easier to pronounce or spell, impossible to lie in, or some such goal. Often they based these goals on a belief in certain universal linguistic principles that their language will display more effectively than any others.
The problem, of course, is that their new languages usually make much more sense to the creator than to anybody else. And the new languages tend to promote schisms among their adopters at a startlingly early stage in the process, often when they have only a few hundred users at most.
And even when there is success with creating languages that are inherently more logical and precise, such as Loglan, a key problem arises: it is just too damn hard to figure out how to say anything in the natural flow of a conversation. People have to think too hard and in somewhat unnatural ways to communicate adequately. In other words, many of these ideal artificial languages assume a clarity of thought and intent that real people cannot live up to in the actual world. It turns out that while many people want to blame the limitations of language for our difficulties in communicating with each other, much of the problem resides within our own minds. "Natural" languages (I don't think the term is entirely accurate, as it seems to me that all spoken human languages are artificial in origin, but it suits how we tend to think about language) work in large part because they are flexible enough to deal with our own fuzzy thinking.
My favorite part of the book came at the very end, when Okrent describes the constructed language or "conlang" community, which consists of people who make up rigorous, linguistically accurate languages for fun, often basing them on elements borrowed from many different obscure terrestrial languages or coming up with really esoteric ideas suited to alien species. I don't think I'd want to do the work that those people put into their hobby, but I can appreciate the creativity and discipline that it takes. Some of the results, such as the squirrel language Dritok, which consists of clicks, pops, hisses, and snorts, sound very entertaining.
An interesting book, though not one I'd read if I was pondering the idea of creating a new language for the betterment of humanity, because the text is littered with the broken sentences of those who tried and failed.