Raghuram Rajan explains it like this: artificially low interest rates are a tax on savers and a subsidy to spenders, and anytime the government distorts market prices it creates unforeseen consequences that can make the situation worse. Here's an excerpt from his article "Money Magic," but read the whole thing.
More than any other policy action, monetary policy suffers from the sense that there is a free lunch to be had. Yet the interest rate is a price for the savings that are transferred to spenders. To the extent that the Fed manages to push this price down ... , it taxes the producers of savings and subsidizes the spenders of savings. Clearly, no government considers pushing down the price of any real good an effective way to stimulate the economy – any gain to consumers is a loss to producers, and the loss typically will outweigh the gain if the market price is a fair one. So why are savings different?
Clearly, someone is paying a price for ultra-low interest rates: the patient and uncomplaining saver. Interestingly, if traditional spenders such as firms and young households are unwilling or unable to take advantage of low interest rates, low rates could even hurt overall spending, because savers like retirees receive lower financial incomes and curtail spending.
I've also argued that an early and unexpected move by the fed to tighten monetary policy (say, by raising the Fed funds rate to at least 1% and announcing that more hikes are coming, much as the ECB has done) would actually help the economy by a) strengthening the dollar, b) punishing speculators, c) encouraging investors, and d) reducing the risk of an inflationary policy mistake, thus boosting confidence. I'm sure Rajan would agree, since higher interest rates also would reduce the current distortions to saving and investment.
HT: Greg Mankiw