BENJAMIN H. FRIEDMAN: Whether or not this president’s vast expansion of war powers is legal, Congress should make it illegal by exercising its war powers to restrain the president further.

That said, it’s worth noting that the absence of legal restraint on the president’s ability to order drone strikes is a consequence of good fortune. Our power, safety, and technological prowess that allow us to fight wars — particularly from the air — at almost no cost to ourselves is a good problem to have. But it’s still a problem. Democracy functions poorly without obvious costs. It prevents conflict and debate, which allows ill-conceived policies to fly under the radar.

Unfortunately, because this problem is a product of our good fortune, safety, and power, there’s no solution to it. But we can improve matters by pushing for these wars to end, resisting the threat inflation that justifies them, and encouraging Congress to more jealously guard its war powers. Ultimately, making the cost of wars more evident will hopefully encourage the formation of more antiwar interests.

Through the authorization of military force that Congress passed in 2001, the president was empowered to use “all necessary and appropriate force against those nations, organizations, or persons that he determines planned, authorized, committed, or aided the terrorist attacks that occurred on September 11, 2001, or harbored such organizations or persons.” In practice, that — along with the constitutionally based self-defense power that the executive branch has — has become a warrant to detain or kill whomever is in their midst across south Asia, the Middle East, North Africa, and beyond.

In theory, it justifies even more. The last two administrations have adopted almost limitless interpretations of the powers this law grants. It’s hard to say exactly how limitless these claims are. The administration refuses to share its reasoning, even its legal reasoning, with most of Congress — let alone those of us in the public.

Besides that, the rules aren’t law. They’re bureaucratic procedure. The administration says that they don’t need any more legal guidance from Congress. They want the courts to stay out of these matters. As such, there’s nothing stopping this administration from changing the rules in secret tomorrow — or for the next president to just ignore them entirely.

That’s not a great situation for democracy —where we can’t even figure out what the rules are governing where we can strike. We don’t know exactly what countries the administration claims it now has the power to strike. But the real culprit for this overly broad and unrestricted war on terror — of which drone strikes are just the most visible part — is not the president. It’s not the lawyers who tell him he can do whatever he wants. It is the Congress that lets him. Congress has abdicated its foreign policy powers.

But it’s difficult to say that Congress ought to do this or that without giving any thought to why there’s been this abdication of power. And there are a variety of explanations for why this has happened.

Many blame partisanship. But to me, the real explanation again is our power and safety, which makes promiscuous use of force abroad costless, or at least seemingly so. Power restrains power. As Thucydides has the Athenians saying in the Peloponnesian War, “The strong do what they will, and the weak suffer what they must.” As the United States, we do what we will, and we invent legal rationales once we’ve done it.

For the Greek city-states that Thucydides was writing about, war was incredibly consequential. If they lost, their city might be overrun, sacked, the women enslaved, and the men killed. But today, even foreign policy disasters like the War in Iraq brought little worse than marginally higher tax rates and unsettling images on television for most of us in the United States. In fact, the volunteer military itself has a norm that says they don’t complain about war in public, which also serves to keep the perceived costs of conflict low. Therefore, foreign policy issues tend to rank low among voter concerns. They contribute little to voting decisions. Therefore, politicians have little incentive to cater to voters’ foreign policy views until costs in every sense start to mount.

Drones simply exacerbate that problem. They lower the costs even more. People enthusiastic about drone strikes like to say that it beats invasion. “Would you rather put troops in there?” they ask. That’s the wrong question, because the alternative to drone strikes should be peace. We’re currently waging bombing campaigns that we otherwise wouldn’t have.

We need to nurse the antiwar and antimilitarism ideology back to health in this country, institutionalizing it now while the time is ripe. More generally, we need to spread the idea that war is something that should be difficult to start.

SPENCER ACKERMAN: Let’s talk about the technology behind these flying robots of death. The conventional wisdom out there is that drones are machines that fly around without pilots. They spy on you according to algorithms that determine where they should go, how they should get there, and for what duration they should stay. And finally, when they see a bad person, they kill them. None of that is really true. These are a bunch of either poorly understood observations or badly articulated experiences.

To begin with, drones are not synonymous with the technology they carry — the sensor packages, the cameras, and the weapons that strike people. Drones are simply airframes, and nearly all of them are in use currently in the military. In fact, as of January 2012, nearly one in every three U.S war planes is a robot.

Nearly all of them have a pilot somewhere. Whether that pilot is a few hundreds of miles away or 7,000 miles away, when you go into the ground control station of a predator or a reaper drone, what you see is something that’s not entirely unfamiliar in the history of manned aviation. You have someone sitting in a dark, freezing cold box, usually with a contractor from the company that manufactures the airframe next to them, with lots of computer screens. There they chat with one another in what looks like Internet Relay Chat, looking over to their chains of command amidst a room littered with lots of energy drink cans due to the fact that they’re on duty for between eight and eleven hours at a time.

Right next to them is a throttle, allowing them to physically control the aircraft in real time. They stay on station as long as the drone is in operation. What that means is that we shouldn’t think of these things as autonomous creatures. There’s a human being controlling them.

That doesn’t mean, however, that new technologies aren’t on the horizon. In May, I was on the deck of the U.S.S. George H. W. Bush for the launch of the X‑47B, a demonstrator aircraft with an enormous 62-foot wingspan. Unlike any drone before it, it’s capable of launching off an aircraft carrier, which is one of the hardest maneuvers in aviation. It’s basically a robotic Top Gun.

The Navy’s upcoming drones, called the Unmanned Carrier Launched Airborne Surveillance and Strike System (UCLASS), will likely be the most advanced drones in the U.S. fleet, entering the Navy, it hopes, sometime between 2018 and 2020. Instead of being controlled by a human being at all times, these drones use lines of code. They run off software programs which, through algorithms and interactions with GPS, program in a flight plan for the drone to follow. When they need it to come back down, another program is entered, and the robot executes it.

Keep in mind, however, that no one in the military is envisioning automating these protocols for weapons strikes. In fact, that is something that the military takes exceptionally seriously and something international law takes even more seriously. Is it technologically possible? Yes. But as of right now, there’s a determination inside the military that this will be crossing a Rubicon, and they don’t want to do that.

The history of the development of this weapons technology, as well as the development of all civilian technology, is the encroachment of automation by very subtle degrees.

One of the things you can do with a robot that can stay aloft for so long, is strap on its belly cameras and sensors of increasing sophistication. The term the military likes to use for this is “persistent surveillance.”

The advanced drones stay overhead for a period of up to 36 hours and stare at what will eventually be called pattern of life activity. It can see, for instance, who goes to the market and when, what their patterns and routines are, and who they interact with. From 5,000 or 10,000 feet — and sometimes as much as 25,000 to 30,000 feet — the cameras are good enough to see down to people’s faces fairly clearly, depending on factors such as cloud cover.

Those are some of the capabilities of these things. What are their weaknesses? In short, these are really terrible aircraft. They fly extremely slowly. They can’t maneuver very well. In fact, the engine used for a prototype of a predator is the descendant of a snow-mobile engine that was put together in someone’s garage. The United States is just starting to put jet engines in them now.

Finally, because many of these drones are GPS reliant, they are very easy to spoof. It’s also very easy to install malware in them. It is somewhat telling that the countries where drones are most often deployed —Afghanistan, Somalia, Yemen, tribal Pakistan, Libya after the initial aerial barrage —are countries that don’t have sophisticated air defense systems.

The drone budget in the U.S. military is declining, meaning that with some exceptions, they’ve bought as many airframes as they need. Where does the drone market go next? It comes home to the United States. The Federal Aviation Administration has said that it will not allow the use of weaponized drones for civilian law enforcement.

But by 2015 the FAA is committed to vastly opening up its licensing for the use of unmanned aerial vehicles in U.S. airspace. Drone manufacturers have a lobby, called the Association for Unmanned Vehicle Systems International , which is working hard to push the rapid acceleration of these drone licenses. Moving forward, this is an issue that will be worth paying attention to.

JULIAN SANCHEZ: The domestic use of drones is probably not going to mean, as Sen. Rand Paul worried, that a Hellfire missile will interrupt your café experience. Nevertheless, it is clear that drones are coming home. The FAA predicts a rapid expansion in the presence of drones over U.S. airspace in the near future, projecting that 10,000 will be in our skies within five years and 30,000 by 2030. More than 220 public agencies have already expressed interest in the use of drones for various applications. And both cities and states are already beginning to contemplate legislation to limit their use.

This is a pattern we’ve seen throughout the history of technology. A platform is developed with military funding for military applications — but then, as war winds down, it finds its way into a variety of civilian and commercial uses. There is a whole array of mostly benign surveillance applications —from search-and-rescue missions, to traffic and weather monitoring, to hostage interventions —that drones can help revolutionize.

But even with these cases in mind, the prospect of widespread drone use raises a number of serious privacy concerns. Perhaps unusually, this is a case in which I recommend a preemptive federal regulation to address these concerns. I believe this is an appropriate course of action for several reasons.

First, if these privacy concerns remain unresolved, it will be difficult to predict which applications of drones will be illegal in the future. As such, both public and private investment may be limited, thereby preventing us from taking advantage of some of the more benign uses of drones. Clarifying the rules in advance is likely to steer investment in more efficient directions.

Second, legally speaking, the extent to which a specific technology is in general public use is a critical factor in determining whether its use is considered a search when it’s in the hands of law enforcement officials. In other words, if there are current civilian uses of drones that make us uncomfortable, it is important to clarify this in advance —before it becomes so common that the court is less likely to impose its own limitations.

Finally, remedies at the local level are not going to be as effective in protecting privacy. To begin with, this is inherently an interstate phenomenon. These are flying objects with extraordinarily powerful sensors — and photons have little reason to respect state boundaries.

A patchwork of state laws for drone privacy is likely to be unwieldy and unworkable. At the private level, tort remedies are going to be less effective in redressing private harms. Why? When dealing with quiet, high-flying objects carrying powerful sensors, identifying and attributing a privacy violation is likely to be extremely difficult.

These factors all suggest preemptive action on a federal level. But as we think about establishing rules around the domestic use of drones, we may want to approach that task by breaking up the issue into different dimensions.

One dimension is the sensors themselves. Taking into account the height at which these drones are expected to fly, it’s important to consider whether the strength of the sensor is proportionate to its use. So, for traffic monitoring, it’s probably unnecessary to have a sensor powerful enough to look through your bedroom window and count the freckles on your back.

Another dimension is duration. If a drone is tasked with monitoring a particular person or group, rather than simply conducting general aerial surveillance, it may be appropriate to establish some period of time after which a warrant would be required to continue such surveillance.

The retention of data is another dimension to consider. The aggregation of information about us — even when we’re in public —as well as the degree to which that information is searchable has direct implications for the level of privacy we enjoy. For example, if drones are used to monitor real-time traffic, it may be important to have certain constraints on the amount of time that information is retained. Without these constraints, we run the potential of creating enormous databases that record public activity over long periods of time — databases that could later be searched.

A final dimension to consider — though there are undoubtedly countless more — is the issue of access. As the Associated Press recently discovered, there is a very low threshold legal standard when it comes to subpoenaing records held by third parties. We may want to silo drone data in some fashion so that it cannot be easily aggregated, thereby imposing additional constraints on government access to that data.

The important thing to remember is that these questions on privacy are in some ways different in kind, rather than degree, from the questions that have emerged from previous technologies. Drones are relatively cheap.

They are small, silent, and have the ability to hover outside of your bedroom window in ways that other aerial surveillance technologies never have. The assumption of reciprocity —your ability to lower your voice, for instance, if you don’t want to broadcast your cell phone conversation in public — is inherently diminished when it comes to drones.

What precise standards should be implemented to address these issues is not entirely clear. But what is clear is that we should think about each of these dimensions and how they interrelate when establishing rules for privacy. And we should do it before the drones get up there.