When I think of lasers, I think of precision optical devices that take hours of careful alignment to optimize. The reason for this is that stimulated emission is usually much weaker than spontaneous emission and without all the feedback created by carefully placed high reflectivity mirrors there would be no laser. This description, while accurate, ignores the overall physics behind the laser, which is best seen when looking at random lasers.
Random lasers have been around for a few years now and a recent article in Journal of Applied Physics has given me an excuse to write about them. Instead of creating your laser from a nice optically transparent medium surrounded by mirrors, you have simply blast a powder with energy, causing the particles to glow. The trick is that each particle in the powder will reflect light in all sorts of directions and between certain pairs of particles there will exist a path for the light to travel back and forth as though the particles were mirrors. Between the two "mirrors" the light will travel along a complex, narrow path that is much greater than the straight line distance between the two. Thus, in contrast to a conventional laser, where a large volume of material must be excited, only a few atoms take part in the laser process so it is relatively easy to get those atoms in the correct state to lase. Since all the atoms are ready to lase, the mirrors at the end points don't need to be all that good. The best thing is that in any powder this occurs for huge numbers of particles so you end up with more than one laser. This paper uses zinc oxide embedded in a polymer matrix, which holds the powder in a fixed formation, making it much easier to work with. Zinc oxide, when sufficiently excited, lases in the ultraviolet region, where normal semiconductor lasers will not operate. Zinc oxide also has the potential, through some trickery using nonlinear optics, to produce a huge range of colors from the ultra violet out to the mid infrared – a spectroscopist's wet dream.
These laser sources are the very antithesis of normal laser development. No care in construction is required and the system is incredibly robust. If you shake the thing up and down a bit – something that will cause an ordinary laser to hemorrhage – all you get is a group of new lasers. Of course, since the laser is random, no one really knows in which direction it will emit. Since only a few particles make up each laser, the power doesn't really scale to very large numbers. However, this isn't so bad since many laser applications don't require high power but would benefit from reduced cost. The biggest problem, however, is getting these things started. At present a huge pulsed laser, the likes of which would not fit in your CD player is required, which is kind of a problem for those thinking about monolithic applications.