Maybe Jacques Derrida, the French dauphin of deconstruction, was right: In the beginning and end was the word. Logos. In war, words matter. Take our drone war, which is not, in point of fact, a war, and involves "drones" only incidentally. And yet the concept of hovering, amoral surveillance machines with missiles attached to them is pretty much the way everyone describes a much different reality.
1. The drone war is not fought primarily with drones. The United States targets members of al Qaeda, al Qaeda affiliates and now, apparently, affiliates of those affiliates, using a comprehensive array of technical intelligence resources, backed up by fighter jets with conventional bombs, submarines that launch missiles, other platforms that launch missiles, and, sometimes, missiles attached to remotely piloted vehicles. The policy is best described as targeted surveillance and killing of the aforementioned groups. In certain areas, it is easier to fly airplanes; in certain places in Pakistan, RPVs launched by Afghanistan will do the job. The munition and vehicle used depends on the target, his location, his importance, and the resources available to the military and CIA at the time.
2. The CIA does not "fly" drones. It "owns" drones, but the Air Force flies them. The Air Force coordinates (and deconflicts) their use through the CIA's Office of Military Affairs, which is run by an Air Force general. The Air Force performs maintenance on them. The Air Force presses the button that releases the missile. There are no CIA civilians piloting remote controlled air vehicles. The Agency has about 40 unmanned aerial vehicles in its worldwide arsenal, about 30 of which are deployed in the Middle East and Africa. Most of these thingies are equipped with sophisticated surveillance gear. A few of them are modified to launch missiles. The Air Force owns many more "lethal" RPVs, but it uses them in the contiguous battlefield of Afghanistan.
3. The targeted killing policy is the best of all worst options for two reasons. One: The United States does not have a coherent and legitimate capture and detention policy. (Thank the CIA torture program, Abu Ghraib, Congress, and the Obama administration's weak efforts to create one.) Two: Human intelligence collection has atrophied to the point where there are not enough people on the ground to facilitate the capture and detention of wanted targets. This means the US over relies on technical intelligence, and on signals intelligence in particular. In Pakistan, it relies on tips from the Army and the ISI. Often, the member of al Qaeda core who's been identified by the ISI is not, in fact, a member of al Qaeda core, but is instead a Pakistani Taliban or militant who is not sufficiently pro-Pakistan. The U.S. has gotten better at vetting these tips, but the policy generally is that it's best not to let the sufficient be the enemy of the reliable. Yemen's government does the same thing. The U.S. MUST rely on allied intelligence services because it cannot rely on its own. So: Bad guys exist. Can't capture 'em. Can't figure out who they are without help. What's the answer? You kill them. If you oppose the policy of targeted killing of al Qaeda operatives, then you ought to support a viable detention system as well as a significant increase in our indigenous human intelligence capacity. Special operations forces and the CIA really would like to capture these guys and interrogate them, because these guys will often give up their comrades. But they can't. So they don't. And the president won't take any chances in letting someone potentially dangerous slip through his grasp.
4. Al Qaeda core has not successfully pulled off a plot against the West since 2005, according to Peter Bergen. Most of the militants targeted by the U.S. in Pakistan today have absolutely no interest in attacking the U.S. homeland. They DO have an interest, a series of very parochial interests, in Pakistan and Afghanistan. At some point, it makes no sense to chase down every person who ever uttered a threat against U.S. troops in Afghanistan. The threat from al Qaeda affiliates (and still the intelligence community bickers over the definition of what an affiliate is) is less than the threat from al Qaeda core.
5. RPVs are NOT the future of warfare. They are a future part OF warfare. Wars are still mostly fought by people in the theater with guns and ammo and communication trucks. RPV technology is advancing, but it is still hard to get one of those buggers to hover in place for an hour and THEN shoot something, and then hover for hours. It's doable, but hard. (Most battle damage assessments are done with other UAVs). That's why the RPVs "orbit." Their courses are programmed; they can deviate off-track and be rapidly reprogrammed, but physics still prevents complete freedom of movement especially if the UAVs have large ordinance on board. If an intelligence source has the exact coordinate of a known al Qaeda operative, the weapon of choice used to kill him will be the platform that is closest, available, and would provide the least collateral damage and most accuracy, depending upon the mission and its own operational security needs.