We have all seen news articles about the FAA's notice of proposed rulemaking to govern the commercial use of drones in the U.S. The FAA's own summary sheet is shorter and better-written than nearly all the articles, so take a look at their Summary of Proposed Changes.
First of all, I am surprised by just how conservative the FAA is being around the issue of pilotless flying aircraft. The category not only limits altitude (below 500 feet above ground level), speed (100 m.p.h.) and weight (25 kg), but more importantly bans operations over non-participating people (football games, unsuspecting crowds, downtowns, etc.). But the really interesting level of care comes from the requirement that the operator always be able to literally see the drone if needed, even if a visual observer is doing the observing for the moment. The operator really cannot be miles away under these rules, and furthermore the operator or visual observer can only care for one drone at a time.
This last rule eliminates the possibility of a commercial operator having a fleet of twenty drones, all overseen by one human supervisor, doing agricultural assay, bridge inspection, disaster mapping, or any other commercial operation you can imagine. The fundamental position implicit in the FAA's rulemaking is that drones are really miniature airplanes that happen to have their pilots on the ground. They are not looking at drones as intelligent machines, but as the flying shell of an intelligent, remote pilot. As such, the drone cannot operate if there is poor visibility; the drone cannot operate if the 'pilot' cannot see it when needed, and the drone cannot share a pilot with other drones at the same time.
Ironically, as technology marches forth drones will become ever more intelligent; and with every advance, they actually become safer, in many cases, than the human pilots can ever be. If the visibility is near zero, it is the well-instrumented drone, with on-board sensor processing and control, that can fly at a disaster site and look for signs of human survivors. In such cases humans are nothing more than supervisors- not pilots at all- and under these circumstances, the "one drone per operator" policy is very much counterproductive and outdated.
So the rules are a good first step if we are to ease very gradually into the world of commercial drone operations: they will be matched one-to-one with operators, and will start out in unpopulated and controlled sites, where bridge inspection, agricultural assay and movie production are all possible applications. These rules are a serious problem for startups and heavyweights who have been dreaming of operations atop our nation's major cities, and they will have to wait patiently until the first flush of drones prove or disprove their usefulness, reliability and worst-case characteristics in time for new rules in several more years. The FAA rules are incremental, and in this case that's exactly the right thing to do.