One of the main criticisms of Bitcoin amongst economists from different schools of thought such as Paul Krugman or George Selgin is that it does not have any value outside of being medium of exchange, so it does not have a backstop on its demand which would underpin its function as a medium of exchange.
Let’s go first through a simple analogy: The first pdf file format was aimed to substitute written physical paper, that was its only purpose. But physical paper on the other hand has other uses. You can make paper fire bricks to use it as fuel, or to prepare many different kinds of handicrafts, or it can be recycled for different purposes. Do pdf files need a different backstop to its demand in order to be useful? Are then pdf files “almost” useless? Should have we refrained from using pdf files until they had more uses or until a more multi-purpose format was invented?….
The problem here is that the marginalist revolution is 150 years old, but economists still confuse value and utility, and very specially when it is related to currency as they require currency to be valuable in order to be useful.
While it is unquestionable that as any economic good has more different uses, chances are that it will have stronger and more stable demand, that is not a reason to qualify as useless or almost useless a thing that only serves for a specific use, no matter it that use is being a medium of exchange. This is applicable to all economic goods, it is nothing special for money.
One thing is to historically observe that money emerged from goods that had non monetary utility (salt, cattle), and one other very different thing is to take that historical observation as a theory. That would be like theorizing that cutting tools must be first discovered through stone and metal and then claim that plastic cutting tools are theoretically impossible without first having gone through the process of being made of stone and metal.
In this respect, Mises Regression Theorem is a very unfortunate interpretation of Menger historical observations. Mises Regression Theorem is wrong or at best completely unnecesary as it was demonstrated by Austrian-Mengerian Carlos Bondone (2006 p.111) before Bitcoin even existed, and as I briefly explain here and here.
The problem that most economists have is that they are still partially anchored on the Theory of Objective Value and do not fully accept or understand the Theory of Subjetive Value. Claims like “but you cannot eat gold or Bitcoin” or “at least you can use gold for jewerly” are reflections of that anchoring to natural sciences in the sense that “if I can not wear it, eat it or touch it, it is not completely real” wrongly assigning some kind of virtuality or artificiality to money. This is a wrong approach.
There are many different human needs apart from the more basic biological and physical needs. Knowledge and institutions such as language, moral (law) or money fulfill needs that are in my opinion much more economically relevant than basic physical needs.
More specifically, the need for exchange or need for liquidity is an extremely important need. It is obvious that without the ability for exchange, humans would be in a much precaurious condition. So anything that might fulfill the need for exchange is useful, no matter if it has other uses or not.
There is no such a thing as intrinsic value, there are only intrinsic properties that might render utility and therefore value. But utility and value are always subjective ideas that reside within the human mind, not within things.
Bitcoin seems to have the intrinsic properties for being money (durable, portable, divisible, relatively scarce, difficult to fake, etc), but Bitcoin’s success as a medium of exchange has to be decided by free markets and we are now privileged to witness this amazing process in real time.