Human Nature: Science, Technology, and Life.

Bush on “Over-the-Wall” Drones

I went to see President Bush's farewell chat with the American Enterprise Institute yesterday. It was an unusually frank conversation: He actually admitted mistakes. In fact, he wandered so far off-message that when he was asked about defense spending, he started talking about specific weapons systems, and pretty soon he said this:

Our soldiers are carrying unbelievably new technologies, using Predators to use over-the-wall intelligence to be able to have better battlefield awareness.

Over-the-wall intelligence? I've searched DefenseLink, and I don't see that term or anything like it. I do, however, see "through-the-wall surveillance," associated with the Air Force Research Laboratory. And the last time I checked, walls tended to be associated with roofs. So one way or another, the Predators have to see through something.

Is Bush possibly blabbing about the technology cryptically described by Bob Woodward and others? The ability of U.S. unmanned aerial vehicles to identify and track human targets "even when they are inside buildings"? If so, the convergence sketched here two months ago—unmanned vehicles that can see through walls—is indeed upon us.

By the way, as of this week's hit in North Waziristan, the number of U.S.  drone missile strikes in Pakistan this year is approaching 30, and the body count is over 200.

Drones Over Dakota

Good news for U.S. national security! We're finally flying surveillance drones over the treacherous border ... between us and Canada.

You're kidding me, right? William Arkin and Peter Feaver of the Washington Post have the same question I do: Why are we posting drones in North Dakota when we need them in Pakistan?

This weekend's national coverage of the first drone's arrival (courtesy of the AP and NYT) didn't really explain. So I moseyed over to the Grand Forks Herald (before I saw the same link at Danger Room, I swear) to see what I could find.

Answer: Pork. Here's the Herald:

At 2:28 p.m. Saturday, the unmanned aircraft touched down at the Grand Forks Air Force Base ... The Saturday landing's significance is about the new mission of the base, which has lost planes and people in large numbers in recent years. The Unmanned Aircraft System mission allows the base to continue, although in a reduced capacity. [Base commander Col. John] Michel said the base will be home to more than 20 UAVs in a few years. The Air Force portion of the 20 will be six Global Hawks, a bigger and higher-flying version of the Predator, by early 2011. "Our manpower will be shored up by 843 people by those Global Hawks," Michel said. "The UAS is the fastest growing part of the (Air Force) business."

How's that for irony? Drones, a technology we're fast-tracking and fast-deploying in the name of replacing personnel, are becoming a jobs program. Though I guess that's consistent with the current economic meltdown and the new emphasis on finding work for people instead of the other way around.

Among other things, it raises the question: What's the future of Air Force bases? How many will fly old-fashioned manned aircraft? How many will be more like NASA's mission-control facilities? That's one of the nice things about drones: You can operate them from anywhere. Which means we should be able to create drone-operation jobs in North Dakota while flying the drones over Pakistan, Afghanistan, or India, not Canada.

Armed Robotry

I've been meaning to get back to this Cornelia Dean piece from last week's NYT Science Times. It's about one of my favorite topics: military robots. Except it confounds some of my assumptions, which makes it all the more worth thinking about.

First off: The "killing machines" I keep writing about are just drones. They're fully controlled (except for malfunctions and weather) by human pilots. Dean is talking about something way more unnerving: machines that make their own killing decisions. I had assumed that for safety reasons, this kind of technology was still confined to the computer equivalent of drawing boards. Wrong. Army software contractor Ronald Arkin tells Dean that armed mechanical border guards are already on the job in Israel and South Korea. Here in the United States, the Army is paying Arkin and others to explore, among other things, how to design such robots to "operate within the bounds imposed by the warfighter." In other words, before we give them guns, we'd better figure out how to keep them from screwing up royally or turning on us.

What's really interesting about Arkin is that he directly contradicts my paranoid prejudice. It's not the armed robots I should worry about. It's the armed humans. Dean summarizes his argument:

In a report to the Army last year, Dr. Arkin described some of the potential benefits of autonomous fighting robots. For one thing, they can be designed without an instinct for self-preservation and, as a result, no tendency to lash out in fear. They can be built without anger or recklessness, Dr. Arkin wrote, and they can be made invulnerable to what he called "the psychological problem of ‘scenario fulfillment,' " which causes people to absorb new information more easily if it agrees with their pre-existing ideas.

His report drew on a 2006 survey by the surgeon general of the Army, which found that fewer than half of soldiers and marines serving in Iraq said that noncombatants should be treated with dignity and respect, and 17 percent said all civilians should be treated as insurgents. More than one-third said torture was acceptable under some conditions, and fewer than half said they would report a colleague for unethical battlefield behavior. Troops who were stressed, angry, anxious or mourning lost colleagues or who had handled dead bodies were more likely to say they had mistreated civilian noncombatants, the survey said.

That makes sense: In war, emotion is more hindrance than help. Same goes for my previous speculation that pilots will become more brutal as they're insulated from physical risk. Arkin's data suggest that in fact, exposure to physical risk makes troops more aggressive, not less. Again, the theory makes sense: You shoot first and ask questions later when failure to shoot jeopardizes your safety. Take the ego out of it—make you a robot instead of a person—and the self-protective instinct to shoot first disappears.

That leaves the problem of ethics. Hormones, mirror neurons, socialization, and love, among other things, make most people reluctant to kill one another. Robots lack these inputs. Will they be ruthless? Arkin's answer, as related by Dean, is that "because rules like the Geneva Conventions are based on humane principles, building them into the machine's mental architecture endows it with a kind of empathy."

Well, I wouldn't go that far. It's not empathy, exactly. But maybe empathy isn't so hot as a guide to behavior in combat. Maybe one lesson of the Army's Iraq survey is that empathy too easily morphs into tribalism. Maybe mechanical soldiers programmed with ethical rules, like the machines of I, Robot, are more likely to behave decently.

But then comes the hitch: What happens when the grainy realities of war defy the simplicity of the robot's program? What happens when the hard part isn't restraining yourself from firing on civilians, but distinguishing them from enemy forces in the first place? That's where Arkin's dream bogs down. He admits it would be hard for robots to recognize physical changes that entail moral changes, such as an enemy fighter with a wound or a white flag. And that's basic stuff compared to the multiplying subtleties of modern counterinsurgency. It's not as though al-Qaida hands out uniforms. Is the guy with the backpack a student or a terrorist? Is the woman across the street chubby or wearing a belt full of explosives?

Here's my preliminary take on Arkin's idea: He's right that we can and should substitute robots for humans in some lethal jobs. Where the categories are clear and cold reason is crucial, let the robots do the guarding and killing. But don't give the early generations of robots any jobs that require nuanced judgments about who's a bad guy and who isn't. And be prepared for the bad guys to learn the loopholes in the robots' algorithms. If the robots respect white flags, the terrorists will use white flags. If the robots presume women are civilians, the terrorists will use women. That's what terrorists do: They study our habits and exploit them. It's a human skill. And it will take humans, not robots, to defeat them.

Droneheads

Several of you have implied in the Fray that I'm obsessed with drones. This is really quite unfair. If you want to see true obsession, visit Wired's military blog, Danger Room, where you can read more than 300 posts about unmanned aerial vehicles, or unmanned aerial systems, or whatever the hell the Pentagon (or is it the CIA?) is calling them now. The DR crew doesn't just follow the drone war in Pakistan from afar, as I do. They go straight to the Pakistani press.

Take a quick stroll through their recent archive and you'll see that they're way ahead of me in tracking the revolution. For example: Pakistan already has spy (not killer) drones. The Russians are buying drones ... from Israel. Iraq wants drones, too.

I could spend all day at that blog. But that might be a tad, you know, obsessive.

Eyes and Ears

A week ago, when we last checked in on the drone war in Pakistan, the news wasn't good. Insurgents had bombed a Pakistani hotel and a security checkpoint, apparently in retaliation for drone strikes on them. The Pakistani government, in turn, was asking the new U.S. commander in Afghanistan, Gen. David Petraeus, to call off the drones. Petraeus said he'd listen. It looked as though the United States might buckle.

Then Petraeus went to Afghanistan and praised the drones. "It is hugely important that three of 20 extremist leaders have been killed in recent months," he told the AP. And on Friday, the Pakistanis got their answer. A drone attack killed another dozen suspected militants at a Taliban commanders' house.

The machines have now racked up more than 100 kills in Pakistan since August. Petraeus has been lobbied, and Barack Obama has been elected, but the drone strikes go on.

How is Pakistan greeting this aggression? Is it threatening to fight? Hardly. Yesterday the country's president told the AP, "We feel that the strikes are an intrusion on our sovereignty, which are not appreciated by the people at large, and the first aspect of this war is to win the hearts and mind of the people."

"Feel"? "Not appreciated"? It's hard to come up with weaker language than that. The real message seems to be: Do what you must, but try not to give us political trouble.

From that standpoint, drones are a lot less harmful than the alternatives. The biggest popular anti-American protests in Pakistan recently were triggered not by drones but by a U.S. ground incursion. Likewise, in Afghanistan, recent politically incendiary mass killings of civilians have been inflicted (accidentally) by human operators on the scene. Yes, the drones have killed some Pakistani civilians. But not nearly as many, it appears, as Pakistani forces have killed in their own clumsy campaign against the insurgents.

Why do the drones have a better record of minimizing mistakes? For one thing, they don't have to make quick decisions. They can hover, watch, and wait. The intelligence they collect can be sifted and weighed by multiple supervisors before reaching a decision to fire. And in Pakistan, they seem to have an additional asset: human sources on the ground. The Washington Post explains:

Brig. Gen. Mahmood Shah, former longtime head of [Pakistani] government security in the tribal areas, said the missile attacks have become noticeably more precise, leading some to believe that local tribesmen in the border areas are supplying the U.S. military with better information about targets. Shah said rumors about so-called U.S. spies among the tribes have fed paranoia about the possibility that signaling devices have been deployed in area villages. Tribesmen have lately made a habit of sweeping the areas around their homes for such devices, he said. "They're not sitting outside in their compounds anymore because they are afraid that they will be struck by these missiles," Shah said.

All this time, I've been looking for technological answers to the mystery of the drones' precision, their increasing ability to find the bad guys. But maybe the answer isn't machines. Maybe it's people.

And if it's people, then the bad guys don't have to fight the machines. They can do what they already know how to do: kill some people and intimidate the rest. That seems to be what they're trying. A day after Friday's drone strike, Agence France-Presse reported:

Taliban militants killed two Afghan men Saturday in Pakistan's restive tribal belt after accusing them of spying for US-led forces. ... The executions were the latest in a string of similar killings and come a day after a suspected US drone fired missiles and destroyed an Al-Qaeda sanctuary in North Waziristan, killing 14. ... Executions routinely follow suspected US missile strikes against militant targets in Pakistan, which officials say are often conducted on intelligence provided by paid local informants.

According to the AP, the two bodies were thrown onto a road, each pinned with a note that said, "See the fate of this man. He was an American spy."

Were the men really spies? If so, were they scouting targets for the drones? I don't know. But for the last three months, somebody's been doing a heck of a job finding the bad guys in northwest Pakistan. Maybe, as U.S. military sources have let on, it's the drones themselves. Or maybe that's the cover story for what's still the world's greatest enemy-detection device: the human being.

Robot Proxy War Update

I can't keep up with the drone war in Pakistan.

This morning, I posted a piece on the evolution of the Pakistan border conflict into the world's first robot proxy war. There have been so many drone strikes along that border in the last four weeks that when I linked to the reports on all of them, it felt like-pardon the reverse metaphor-overkill.

Now it turns out I missed one. The machines' body count is now 20 higher, thanks to a strike last night. It's the 19th drone attack since August. According to an update this morning on the New York Times Web site, the strike occurred 20 miles inside Pakistan and took out two Taliban commanders who have launched raids on U.S. troops in Afghanistan.

How good are the drones? According to the Times, one of the targeted commanders "was believed to have been visiting the compound ... to pay his respects to the families of those killed in an American drone strike on Friday" in a different location. The machines find and kill you, and then, when your boss shows up somewhere else to console your relatives, the machines are waiting for him there, too.

Down the road, we should all be scared of what this technology can do. But for now, I'm enjoying our ability to find and kill these guys without putting boots on the ground.

Now, about those other 18 casualties ...

Evolving Predators

Is this technology for real? If so, what is it? Since the government isn't telling, I poked around a bit and asked readers for ideas.

Here are some possible leads. First, Slate reader mark_925 flags a list posted Friday on Aviation Week's Ares blog. The list includes several technologies that have improved U.S. efficiency at hunting and killing adversaries in Iraq. They include:

1. Communications intercept sensors "so sensitive that they can pick up the low-power emissions of handheld cell phones."
2. A targeting system called NCCT, which "instantaneously links the intelligence taken from several aircraft, ships or UAVs at once to locate, identify and target electronic emissions, including communications, and associate them with air, ground and sea radar targets."
3. An "IDM communications module" that links communications signals to visible sources, such as cars.
4. Software that facilitates "change detection" from spy aircraft.
5. Helmet sights that immediately translate a physically viewed object into spatial coordinates that enable fast targeting and destruction.

Second: Walter Pincus of the Washington Post flags an article in the U.S. military journal Joint Force Quarterly, written by the general who, as of today, is replacing David Petraeus as commander of multinational forces in Iraq. It credits the upturn in Iraq in part to a "surge of ... full motion video (FMV) assets." Early in the war, "Commanders were rarely allocated more than an hour of FMV a week," says the article. "Since 2003-2004, FMV within the corps has increased tenfold. ... Today, the corps can count on daily support from at least 12 FMV systems," and each brigade combat team "has an organic tactical UAV platoon that provides 18 hours of FMV coverage a day." Drones are dramatically improving military performance, not by doing the killing themselves, but by providing instant, on-demand customized intelligence to ground forces.

Third: Pincus reports that last week, a Senate subcommittee appropriated $750 million for "intelligence, surveillance and reconnaissance initiatives." This compounds a $1.3 billion shift of money to ISR programs, approved by the Pentagon in July. According to Defense News, the programs include:

1. "$262.6 million to buy digital data links for Raven UAVs, data links and laser designators for Hunter unmanned aircraft, and various improvements for other unmanned aircraft."
2. "$168.5 million to buy eight Medium Altitude Reconnaissance and Surveillance airborne systems, including with $52 million for three new Constant Hawk airborne surveillance and target acquisition systems."
3. "$17 million to extend a contract for Scan Eagle UAV services, $15 million to buy a new Northrop Grumman-made Global Hawk UAV and associated gear and services, $26 million to purchase four Boeing-made Scan Eagles."
4. Imaging systems and "sensor packages" for the Air Force.

So there are some possible clues to the recent turnaround in Iraq and the more recent escalation in Afghanistan and Pakistan: more UAV deployment and video, faster integration of UAV data into ground operations, more acute communications sensors, and instant targeting data on visible objects. Some combination of these technologies might account for the key breakthrough attributed to the devices now being deployed to Afghanistan: nonvisual identification and tracking of targets. Or not.

Over to you, Danger Room.

Terminator 2: Attack of the Drones

Speaking of Terminators: The drone war over Pakistan is escalating.

Boom. Sept. 4: Seven people killed in a strike on Chaar Kehl, near the Afghan border.

Boom. Sept. 5: Six to 12 more killed in a hit on Al Must.

Boom. Sept. 8: 23 dead and at least 18 wounded in a five-missile barrage on Daande Darpkhel.

Boom. Sept. 12: Twelve more dead in an attack on Tole Khel.

That's about 50 fatalities in four strikes in a single week, all at the hands of unmanned vehicles. An impressive warning from the bloodless killers of tomorrow. Even before the hit on Tole Khel, the Washington Post reported, "The number of Hellfire missile attacks by Predators in Pakistan has more than tripled, with 11 strikes reported by Pakistani officials this year compared with three in 2007." According to the Wall Street Journal, "One official in Afghanistan estimated that drone usage in Pakistan has doubled since the summer, and he said missiles are now being fired at Pakistani targets virtually every day."

Why the increase? Media reports from the ground and military sources indicate several factors: 1) Pakistan isn't really helping, so we've taken the killing into our own hands. 2) We don't want to literally use our own hands, since our ground forces might be captured. So, where possible, we're using drones instead. 3) Drone attacks cause less friction with Pakistan than ground incursions do, since U.S. personnel are never at the scene. 4) We're sick of our troops being picked off in Afghanistan, so we're using drones to even the score. 5) We're relying more on drones to spy in Pakistan because we've failed to develop informants on the ground. 6) Or maybe we're getting better ground intelligence, which is giving us more hot targets to shoot at.

The most intriguing factor, however, seems to be an upgrade in drone technology. In Friday's Los Angeles Times, Greg Miller and Julian Barnes report that Predator drones "above the tribal belt along Afghanistan's eastern border" are now "equipped with sophisticated new surveillance systems." The new systems permit

the tracking of human targets even when they are inside buildings or otherwise hidden from Predator surveillance cameras. Equally important, officials said, the systems have significantly speeded up decisions on when to strike. The technology gives remote pilots a means beyond images from the Predator's lens of confirming a target's identity and precise location. ... The technology allows suspects to be identified quickly. "All I have to do is point the sensor at him," said a military officer familiar with the system, "and a missile can be off the rail in seconds." The devices are roughly the size of an automobile battery, but are heavy enough that outfitted Predators in some cases carry only one Hellfire missile instead of two.

Tracking invisible targets? Nonvisual identification? Miller and Barnes don't explain how the system works. All they disclose is that it "was developed as part of a special project within the CIA's Directorate of Science and Technology." But if U.S. drone managers are willing to shed 50 percent of their missiles to make room for these target trackers, they must be pretty valuable.

The arrival of these devices in Afghanistan is only half the story. The other half is where they're coming from. They've been "instrumental in crippling the insurgency in Iraq," according to Miller and Barnes:

A military official familiar with the systems said they had a profound effect, both militarily and psychologically, on the Sunni Arab insurgency in Iraq. "It is like they are living with a red dot on their head," said a former U.S. military official familiar with the technology who, like others, spoke on condition of anonymity because it has been secret. ... Officials said introduction of the devices coincided with the 2007 U.S. troop buildup in Iraq, and was an important, but hitherto unknown, factor in the subsequent drop in violence in that country.

How much of the credit we've given to the troop surge in Iraq actually belongs to these devices? Are they working some similar magic now in Afghanistan and Pakistan? And, if so, what the heck are they? I don't know, and the U.S. government doesn't want to tell us, but I'll keep looking for answers with my primitive human eyes. In the meantime, if you've got any good intel on them, let's hear it.

Offed With Your Head

The Human Nature article on Slate's cover today is about a military drone-piloting system that looks like a video game but kills real people. You control it with joysticks and buttons. The company that developed it, Raytheon, sees it as a logical progression for recruits who come into the military knowing how to play games like Doom and Halo.

The question is: Will the transition be too smooth? Will these young pilots, reclining comfortably in their "virtual cockpits" in Nevada as their drones fly over Iraq, feel as though they're playing a game?

Now imagine taking this merger of games and killing one step further. Imagine controlling the drone directly with your mind. Imagine firing the missile just by thinking it.

Imagination is a dangerous thing. It can already fire weapons in video games. Here's the report from this weekend's Sunday Telegraph:

British scientists are turning the vision into reality with a device that allows objects to be manipulated with brain waves. The prototype ... can already be used to play simple computer games. By imagining a movement, the wearer of the hat-shaped device can tell the computer to move an object around a screen or a robot around a room. ...
The development came as the video games maker Nintendo disclosed that it wanted to build on the success of the motion-sensitive technology used in the best-selling console, the Wii, by developing games that can be controlled by thought.
To pick up the signal from the brain, the scientists use a cap fitted with electrodes that detect changes in the electrical activity produced by the neurons. When a person wearing the cap imagines a particular action, such as moving a hand, it produces a distinct pattern of signals that a computer learns to recognise.

While Nintendo works on deploying this technology, two other companies are already there, according to the New York Times:

Put on the headset, made by Emotiv Systems in San Francisco, and when a giant boulder blocks the path in a game you are playing, you can levitate it—not by something as crude as a keystroke, but just by concentrating on raising it, said Tan Le, Emotiv's president. The headset captures electrical signals when you concentrate; then the computer processes these signals and pairs a screen action with them ... Emotiv plans to have its noninvasive, wireless EPOC headset ($299) on sale in time for Christmas, Ms. Le said. ... So far, [Emotiv's R&D manager] said, all 200 testers of the headset had indeed been able to move on-screen objects mentally.
Another headset, the Neural Impulse Actuator ($169), just released by the OCZ Technology Group in Sunnyvale, Calif., has three sensors in a headband that pick up electrical activity primarily from muscles and convert it into commands ... Players of shooting games, for instance, may use eye movement to trigger a shot, shaving milliseconds off of their response time and sparing their hands.

Scientific American has more on how the Emotiv headset reads your mind.

So now we're looking at two mergers: mind-controlled action with video games, and video games with killing. Firing weapons with your mind used to be imaginary. Now, like so many imagined things, it's becoming real.

I, LawnBott

If I had a nickel for every time I've read the word "robot" in a headline about new technology, I'd ... well, given the current price of metals, I'd melt down all those nickels, sell the ingredients, and become a very rich man. Journalists and PR people use the word "robot" to mean anything from HAL to a remote-controlled toy car. Actually, robots come in various degrees. The revolution we're seeing in mechanization isn't so much in the proliferation of robots as in their increasing autonomy.

Case in point: Two stories from this morning's news batch.

First we have an AP story about a "Bum Bot" designed to disperse vagrants from an iffy neighborhood in Atlanta. It belongs to Rufus Terrill, a local bar owner and ex-Marine. The story says Terrill used to patrol the area on foot, but "guns were stuck in his face several times. His wife suggested he patrol a safer way - using a robot." In the AP photo, the robot looks like a small tank, about half as tall as Terrill. It weighs 300 pounds and has a camera and water cannon. (Terrill says he's never used the cannon.) The robot's exterior has been "nicked by rocks, bricks and other objects people Terrill was rousting have thrown at it."

The point of the robot, it seems, is to take the physical risks formerly taken by its human owner. Any guns that might previously have been stuck in his face now have to be pointed at his tank instead, which doesn't have quite the same effect. There's no report of the tank having been shot, but, as the story says, it has taken its share of rocks and bricks. That's fine. It's part of the plan. Sticks and stones may break my drones, but they can't hurt me.

In this way, the Bum Bot is a lot like the thousands of drones currently deployed by the U.S. military. The enemy can't kill American soldiers who patrol war zones from a safe distance via remotely-operated unmanned vehicles.

The tricky thing about drones, as I've noted before, is that they can desensitize you to the battlefield. I mean literally desensitize you: Your physical senses have no direct contact with what you're looking or shooting at. Can the same thing happen to civilians who use private security drones at home? Apparently so. "It's just like a video game," says Terrill, describing how he operates the Bum Bot. The Atlanta police warn that he might be prosecuted if he uses the water cannon. But there's no such constraint on the use of drones in Afghanistan, Pakistan, or Iraq.

The chief constraint on the Bum Bot isn't legal or moral. It's technological. The Bum Bot isn't really a robot. It's controlled entirely by the handheld remote, and it has no voice other than Terrill's, which he projects through an integrated walkie-talkie. Without constant human direction, the machine does nothing.

If you want to get closer to the cutting edge of robotics, so to speak, you're better off looking at a technology that's already well-commercialized: robotic lawn mowers. Today's New York Times salutes a new product, the Kyodo LawnBott LB3500, which mows your lawn by itself. Here's the manufacturer's description:

It operates automatically, and autonomously by means of its intelligent computer and a perimeter cable. It can move freely within an enclosed area, detecting the faint signal transmitted by the perimeter cable located on the ground, defining the areas to be mowed; it can also work without a perimeter cable as working area is enclosed by a fence or small border at least 4 inches tall. ... [I]t leaves its docking station and starts mowing your yard in a random direction. It will mow in a straight line until it bumps into an obstacle, such as a tree or flower pot, or until it runs over its perimeter cable, then it stops, backs up, turns and takes off again.

Well, at least it needs a human to recharge its batteries, right?

Wrong. The company explains:

When the batteries start running low, or at the end of its cutting cycle, the mower will search out the perimeter cable and follow it back to its docking station to recharge. After charging, it heads back out on its own! ... With the new LawnBott, you have One Less Thing to Worry About.

Well, yes. But you also have one more thing to worry about: Your lawnmower running amok while you're at the office. No human hassle means no human control.

Kyodo says the LB3500 comes with enhanced safety features: "a higher sensitivity, free-floating, 360° bumper shell, blade stop proximity sensor, and an on-board alarm system should an unauthorized user pick up the Lawnbott." Still, we're talking about a slicing machine that runs around by itself and can't even be stopped by power depletion. LawnBotts.com points out that "robotic lawnmowers are many times safer than its manual counterparts just because you eliminate the human needing to be around it while it's operating." This is the same sense in which military drones are safer than manned vehicles and weapons: They protect their owners. But if you're not the owner, look out.

In the AP story about Terrill's bar, some of the locals complain that the Bum Bot is "intimidating." They have no idea what's coming.