Bitcoin will be more interesting to me once the mining pool is exhausted. At that point, we'll see how much of Bitcoin's value is in use instead of speculation.
(I'm not the kind of person to complain about being downvoted, because I literally don't care, but I'd like to know what's wrong in what I said. I always thought printing money is something no sane government would do.)
If the amount of currency is fixed, but the economy grows, then the bread becomes cheaper. This has multiple effects, but one is to value work yesterday more than work today (work today will be compensated more poorly). Another is that it favours people who hold currency already over those that are actually producing useful economic output right now.
So in a growing economy, the currency supply probably should increase. It's arguable that tilting things the other way - favouring work today over work yesterday (or ten years ago) and favouring economic activity over holding cash - is desirable, so most developed economies aim for low but positive inflation.
This is my very simplistic understanding of why completely limited currencies are not a great plan.
First, the marginal utility of additional units drops for the hoarders, and marginal utility is a well established economic fact.
Second is the time preference of goods. People's lives are finite and they don't want money for money's sake, they want to exchange it for useful things. If you're really hungry, you will buy your sandwich today, not tomorrow even if it's going to be cheaper. Real world examples are electronics and the price of oil. Electronics get better and cheaper with time, and yet, people still buy today. Oil has depreciated hugely in the last months, and yet, people are buying gasoline and heating oil right now, even though signs point to it being even cheaper.
Taken together, you have the desire to hoard counterbalanced by these two factors, which creates a balance of hoarding and spending. This does an economy make.
In either case, you're obfuscating the consumer's real demand due to 2nd order speculation. Using some particular good as a unit of account fundamentally introduces a distortion into the way that people elect to spend / save. And many people have different takeaways from this fact. If you're a goldbug or a Bitcoin enthusiast, you think that inflationary pressure is evil and deflationary pressure is good. If you're really into Keynesianism, you might think that consumption is good and saving is counterproductive. If you're the Federal Reserve, you think there is a right amount of spending that expert economists should target by tinkering with the money supply.
For anyone interested, I personally think the real answer is to look for ways to design a system that removes the distortion entirely by introducing a currency that cannot be held. In other words, a financial system in which the unit of account, the grease in the gears of the economy, only exists in the brief context of a transaction. The actual holding of wealth would all be done using electronic "shares" of real material goods, sort of like what you're buying at a commodities exchange. In this world, people's personal savings would be electronic, hyper-diversified stock portfolios. The "currency" of this system, if you could even call it that, simply acts as a yardstick for understanding relative costs, rather than needing to understand the N^2 different exchange rates of a typical barter system. You would hold micro shares in thousands of different products thanks to automatic software tools that blended expedience with your desire to personally elect what goods you wanted a long position on. In this way, the appreciation / depreciation of your personal savings would rely quite transparently on current values of the goods it represented. Crucially, removing this layer of abstraction would make it much harder for your fortune to evaporate purely on perception of value (see: Zimbabwe) since you would never give it away for less than what the underlying goods were worth to you personally.
But I'm utterly unconvinced that 'prices can only increase' for the economy to work. For example, in electronics prices for hardware have generally fallen in nominal and real terms and yet it's still a pretty big industry.
Think of a house: would you buy one if you knew that just by waiting a year its price would be 80% of what it is now? And the year after 65%? You'd probably just wait. If you waited long enough, you could buy one for, literally, today's peanuts.
Electronics sounds like a good counterexample, but electronics are also developing at extremely high speed. The price of a given device decreases very rapidly in time, but new more powerful devices come on the market every day. Sure, you could have waited five years to buy an iPhone 1, but that piece of electronics is now almost worthless compared to the other options you have on the market.