Bruce Trinque's review said much of it, but here is more support. Ruddiman's work offers possible hypotheses to explain many puzzling effects. It is clearly written, accessible to non-experts, and of my 3-feet-wide bookshelf on climate issues, if somebody wanted one book, this is what I would recommend.
Ruddiman offers two basic hypotheses. The first, as Bruce described, is that humans have been modifying climate for 8000 years via forest-clearing and agriculture. This inhibited the otherwise-natural temperature decline back into an overdue glaciation, as compared with past inter-glacial periods. That's the good news.
The bad news, of course, is the current warming that will take us to levels of CO2 and temperature unprecedented for millions of years, and will do so even if we all stopped using oil/gas/coal tomorrow, and he discusses why.
The second hypothesis is the most plausible explanation I've seen for some of the puzzling short-term temperature/CO2 gyrations of the last 2000 years. He proposes that major plague pandemics have caused sufficient die-offs, abandoment of farms, and reforestration to temporarily lower CO2 and temperature. This could explain the later-Roman/Dark Ages lower temperatures, followed by the relatively disease-free Medieval Warming Period, in which Greenland was settled, and UK vineyards spread again to current levels, if not quite as far as early Roman. He ascribes the Little Ice Age drop to Bubonic plagues in Europe, and especially, to the death of estimated 50 million native Americans from smallpox and other European diseases. He does enough math to make these claims at least worth further study. He carefully observes that "correlation is not causation" and goes on to calibrate the mechanisms by which pandemic can lead to lower CO2.
Ruddiman refreshingly understands the differences between early hypotheses and well-tested theories. He often starts with an observed behavior, then carefully evaluates alternate explanations for it, rather than just offering an answer.
This is an exemplary approach to science, and while the hypotheses certainly need testing, this seems like a very productive line of thought that should incite useful further research. Climate analysis always faces the serious problem of extracting trends, and their causes from a very noisy signal. Compared to many competing hypotheses, Ruddiman's seem to be able to explain some gyrations that have often caused people to say "temperatures go up and down randomly anyway."
Finally, the book is clearly and calmly written, with careful delineation of facts and conjectures, with plenty of backup. Even more technical details can be found in his earlier articles.
Finally, I suspect Edward Tufte would be gladded by many of the charts, which are often simple, but compelling, and far better than words.