If nanotechnology involved a smooth transition from the microscale to the nanoscale, it wouldn't be all that interesting. Much of our available knowledge base could be applied, and the possibility of radical new discoveries would be rather small. Luckily (or unluckily depending on your perspective) what happens at the nanoscale is very hard to predict.
That the transition to the nanoscale should be so unpredictable is rather unsettling. From an early age we learn that the properties of materials don't depend on their size. Materials can be scaled up or down readily. A block of wood a certain size can support a certain amount of weight- a larger block can support more weight. Or we learn that a wire that conducts electricity also generates a magnetic field. A larger or smaller piece of metal still generates a magnetic field when current passes through- this property doesn't change with the size of the piece of metal. But using nanotechnology we can have materials that conduct electricity without generating a magnetic field- pretty useful when you want a pacemaker that can go into an MRI.
As an example, a gold nanoparticle does not behave the same way as a grain of gold, a gold nugget, or a gold bar, it has some very different properties. In aqueous solution, gold nanoparticles can be a range of colors, but are most commonly red. In contrast, gold bricks don't go into solution in water. The fact that bulk gold was pretty impervious to most chemicals made it a useful currency, since gold would always be gold. Dunk it in seawater and nothing would happen to it for thousands of years. Gold nanoparticles, on the other hand, are much more reactive than bulk gold, to the point where gold nanoparticles make useful catalysts. Conventionally sized gold is generally not a useful catalyst (there are a few exceptions), but the cost of the material has little to do with its lack of desirability in most chemical reactions. In short- gold nanoparticles have very different properties than the gold we are familiar with, and these properties are impossible to predict from the bulk material. This is because nanoscale materials can also be thought of as quantum materials, whereas bulk materials are considered "classically Newtonian."
At the end of the 19th century, it was discovered that Newtonian physics couldn't explain certain phenomena. While in theory light should be infinitely divisible, in practice it wasn't. Energy in light traveled in distinct packets termed quanta. You couldn't divide a quanta of light any further- that was it. To human beings who perceive things in a Newtonian fashion (we think we can always cut a cake into ever smaller pieces) this concept of quanta was unsettling. To paraphrase Heisenberg (who came up with the oft quoted and even more often misunderstood uncertainty principle) - "Anyone who thinks they understand quantum mechanics the first time it's presented has not comprehended what they have just heard." While quantum mechanics was certainly important to physicists in the first decades of the twentieth century, it probably meant little to the average person on the street. The relevance of this theory to practice was demonstrated most impressively with the atomic bombs that ended WWII. While these atom bombs didn't involve nanotechnology, they certainly showed how quantum mechanics could lead to revolutionary technology.
Unfortunately, there was no way to really manipulate solid materials on a quantum scale, and much of quantum theory remained just that- theory. Theory can only get so far ahead of practical experiment, and without quantum materials to experiment with, quantum theory became largely academic to scientists and engineers working with materials. Another definition of Nanotechnology: the production of quantum materials in significant quantity to solve real world problems