Synopsis
Book by Bratton William Tumin Zachary
Extrait
1
The Case for Collaboration
THE HUNT FOR TEN RED BALLOONS
On October 29, 2009, the Defense Advanced Research Projects Agency (DARPA) announced its "Network Challenge." At 10:00 a.m. on December 5, 2009, at ten locations throughout the United States, DARPA would let fly an eight-foot-diameter red weather balloon tethered to the ground. Each balloon would be readily visible from local roads and buildings--points the average person could reach. A $40,000 prize would go to the first team to accurately report the location of all ten weather balloons.
The contest was meant to replicate the challenge of trying to gather information about an adversary in an open environment. DARPA wanted to test whether ordinary folks using commonly available off-the-shelf technology and social media like Twitter or Facebook could work together--collaborate--to solve a problem that would be, in the words of one expert from the National Geospatial-Intelligence Agency, "impossible to solve by traditional intelligence gathering methods."
A team from MIT's Media Lab won. No surprise there. MIT had a slew of faculty and top graduate students, the most sophisticated equipment, and great publicity. CNN profiled them and drew attention to their cause. A Georgia Tech team placed second, for similar reasons.
Both teams competed fiercely. They put out misinformation, reporting false sightings, sent others on wild-goose chases, and bought time for themselves. Both teams wrote complex computer programs to defend themselves against such attacks.
Given their advantages, you would expect MIT and Georgia Tech to come out ahead--and they did, with a winning time under nine hours.
But what is interesting is the guy who finished in a tie for third with eight balloons, and actually led the pack for the first four hours of the competition--nineteen-year-old hacker George Hotz. Hotz heard about the contest only a couple of days before, and only an hour before it started he put up a website called Dudeitsaballoon.com.
How did he do it? His idea was based on a kind of mass collaboration.
Hotz had nearly fifty thousand followers on Twitter. They, in turn, had hundreds of thousands of followers. His plan was to mobilize them all--get thousands in the game and all those eyeballs searching for the prized red balloons. It almost worked.
Hotz was already famous in the hacker community for "jailbreaking" the Sony PlayStation and the Apple iPhone. He'd cracked their proprietary codes, and for the iPhone wrote software that let iPhone owners use it on any wireless network, not just AT&T's--much to AT&T's and Apple's chagrin and the hacker community's glee.
These legendary hacks made Hotz a star. He gained tens of thousands of Twitter followers, all of whom wanted to be the first to know what George Hotz might do next. On Twitter, they would soon find out.
On the day before the DARPA contest, Hotz--who went by his Twitter name, @geohot--tweeted his followers to stand by for a major announcement the next day. That started a buzz going in the Twitterverse and on hacker bulletin boards.
On Saturday morning @geohot tweeted his fifty thousand followers:
10AM EST today marks the start of a US wide scavenger hunt, for 10 red balloons http://bit.ly/7chum5 #dudeitsaballoon
He quickly followed up with another tweet:
So I need your help to do two things, 1, find big red balloons, and 2, RT [retweet] and trend this !!!! http://bit.ly/7chum5 #dudeitsaballoon
He included a link to his website. The hashtagged #dudeitsaballoon guaranteed that if his message got retweeted, as requested, #dudeitsaballoon would rise to the top of the Twitter trending terms. That would amplify its effect--and call further attention to Hotz's cause.
Visitors clicking through to Hotz's website found the following message:
Right now you are all probably waking up to another normal Saturday. But this Saturday is not normal. In addition to planes, birds, owls, and everything else in the sky, there are 10 red balloons scattered around the United States. Starting at 10AM EST, your US government is using tax dollars to send 10 big red weather balloons into the sky. I need to know the location of those balloons.
So if you see a big red balloon in the sky, about 8ft round, numbered 1 to 10 . . . report it here ASAP so I can win the contest.
Hotz offered $1,000 to anyone who gave him a confirmed sighting. And he offered something that would incite any die-hard hacker.
"Seriously," Hotz wrote. "If you guys come through for me . . . I'll make you an untethered jailbreak."
"Chop chop."
Offering an untethered jailbreak to the hacker community was like dangling red meat in front of a lion. It was the gold standard of all hacks. Unlike Hotz's earlier iPhone hack, which left the iPhone tethered to software you had to run each time you started the phone, this time Hotz was promising to hack the iPhone again and create an untethered jailbreak. Untethered, you could use your phone just like any cell phone, on any carrier. Untethered, the iPhone would be released from its earthly moorings. It would be hacker heaven.
Word raced around hacker online sites and bulletin boards that George Hotz was offering to do an untethered jailbreak for spotting the red balloons. We have to win this, the hacker community buzzed. Do it for @geohot; do it for us!
By hour four, Hotz had four verified sightings--more than the MIT team and the Georgia Tech team. He traded two of his four sightings with one of the other front-running teams. That made six.
Eventually, the MIT and Georgia Tech teams surged ahead, but not before Hotz found eight of the ten balloons. He had done better than dozens of teams competing. It was far more than what traditional intelligence gathering could accomplish.
More than that, it showed DARPA the raw power of the Internet to foster collaboration. What George Hotz lacked in funding, institutional support, and educational credentials he made up for with digital age assets: networks of followers who, on an otherwise ordinary Saturday and with a promise of glory and gifts, he could get in the game fast. Already arrayed on trusted platforms, Hotz sent current through those networks, turned followers into partisans, and got them collaborating--in minutes. Together, they pulled off something extraordinary (and nearly won the Challenge).
RESTORING AN EMPIRE STATE OF MIND
BILL BRATTON Takes New York
As the commissioner of the New York City Police Department and the Boston Police Department, and chief of the Los Angeles Police Department, I learned about the power of collaboration across departments, agencies, and private industry early on.
In 1993, an army of squeegee people seemed to have taken over New York. At every corner and tunnel entrance in the city, you'd stop for a light and they would pounce, some filthy rag or sponge coming up to your windshield, a face and hand close behind. You could try to wave them off. Or you could try to ignore them, eyes straight ahead. Not always practical. It was sort of a mini-street corner protection racket, with the convenient charade of a spit-enhanced wipe down and a key scratch across your car's paint job if you didn't pay them for the "cleaning."
In 1993, the election for New York City mayor was on. US Attorney Rudy Giuliani was running against Mayor David Dinkins, crime, disorder--and squeegee people. You could almost make a compound noun of those terms, lumping them all together, and many voters did. The news stories were incessant, fueling what every New Yorker sensed anyway, whether they commuted by car, foot, or subway: the city was out of control.
Ten percent of New Yorkers experienced violent crime in a year. But every day 100 percent experienced the city's disorder: fare beaters and drunks on the subways, mental patients off their meds wandering the streets, prostitution operating out in the open. Long lines, high taxes, poor service. Broken neighborhoods, broken people, broken windows--a broken city.
It all fueled a sense of chaos. The New York Post summed it up for the incumbent, Mayor Dinkins: "Dave, Do Something!"
Too late for his mayoralty, Dinkins raised money for six thousand more cops. Too often, NYPD commissioner Ray Kelly's cops scattered the squeegee people only to see them rally to some other corner moments later.
When the dust of the November elections settled, the voters had replaced Dinkins with Giuliani; the new mayor soon replaced Kelly with me as NYPD commissioner. I had been the commissioner of the Boston Police Department and before that, in 1991, chief of the New York City Transit Police Department.
Giuliani had made a campaign promise to get rid of the squeegee guys, so I knew I needed to move quickly, continuing the work Kelly had begun. Counting heads, it turned out that the "army" of squeegee men had actually numbered about seventy-five. Well before the Internet, the blogosphere, or the Twitterverse, New York's potent tabloids had turned seventy-five sponge-and-bucket guys into a national symbol of impotent government and a city on the brink.
Persistent police work paid off. Many of the men had had prior problems with the law and couldn't afford to get arrested again. Which is exactly what we promised, and did. We stayed around long enough to break up this thriving little extortion racket that was driving the city crazy. Seemingly overnight the squeegee men were gone--though we did have in our favor thirty-eight thousand cops versus seventy-five squeegee pests.
The tactics I used to conquer that problem formed the strategy of what I hoped would be a much more ambitious effort, one aimed not just at cutting crime but at dramatically changing the quality of life in New York.
The NYPD had people bluffed, as I later wrote in my first book looking back at the time. They had the reputation as the greatest crime-fighting machine in the history of policing, but to me the big blue wall was a lot of blue smoke and a few mirrors.
They were good at responding to crime, they just weren't very good at preventing it. They weren't even trying to prevent it. They were just cleaning up around it.
The NYPD, like many departments, was "all response, all the time." The 911 dispatch system created in the 1970s had democratized policing: it was no longer "who you knew downtown." Now, any citizen could mobilize the department with a free call from a pay phone. And millions did. Police were racing across the city from call to call.
But the 911 system didn't dent crime much--the onslaught of crack, disorder, and guns in the 1980s and '90s saw to that. A single citizen could make hundreds--even thousands--of calls complaining about nuisance gangs, drugs, and prostitutes on the same corner. Officers responded every time, but nothing changed. It was like shoveling sand against the tide--the tide kept coming back.
Remember the precinct house nicknames of the time--"Fort Apache, the Bronx" or "Little House on the Prairie"? That's what American policing had become: isolated outposts, controlling little outside its four walls--or outside the cruiser. The 911 dispatch kept cops in cars, windows rolled up, AC blasting, racing to calls or on "random" patrol in between, intending to deter crime by their mere presence.
As New York City's police commissioner, I quickly set out to establish a new form of policing, one that required collaboration not only between all areas of the department, but also with other agencies and the public. My goal was to transform the city and the American police profession.
It all starts with a vision, I told the department: as good as we are, we can do better. But we can't do it alone.
The path forward--the new platform for policing New York--came to be known as CompStat.
"When have you guys ever addressed crime?" Jack Maple, my right hand at the New York Transit Police Department and now at the NYPD, was digging in. John Timoney, a twenty-five-year NYPD veteran and now my chief of department, had called Maple out for his comment to a reporter. "Those guys over there at the NYPD have given up on crime fighting," Maple had said.
Timoney pointed to this operation and that, and cited his stellar service as commander of New York's 5th Precinct on the Lower East Side. Maple would have none of it. "Your Narcotics Bureau works nine to five, Monday through Friday. The Warrant squad is off weekends. Auto crimes, off weekends. Robbery squad, off weekends. The whole place takes weekends and nights--just when the criminal element gets down to work."
And that was the problem.
To transform the city, I knew, my team and I would have to start with the NYPD. To succeed, I needed believers and doers. I screened the incoming command staff and promoted my own leaders over the heads of others--Timoney among them, and Louis Anemone, who would be chief of patrol. My inner staff was made up of longtime NYPD partisans--but commanders who were loyal to me, who understood and bought into my vision: the NYPD could do better, and this was the way.
Maple had been through this before with me when years earlier I reorganized the New York Transit Police Department. Metropolitan Transit Authority president David Gunn had told me at the time that fare beating was bleeding the MTA dry; disorder was shrinking ridership. There was brand-new capital waiting to be poured into rebuilding the subways--but the subways were out of control. He needed them tamed.
I concentrated patrols where the problem was highest, and ran high-visibility mass arrests. We were able to bring fare beating on the subways down from 170,000 per day to the point where it fell so low that the MTA stopped tracking it. Malcolm Gladwell wrote about this "tipping point" phenomenon in his book of the same name.
But I also learned something that stuck with me: many fare beaters tend to have character flaws. One in seven was wanted on a warrant or probation and parole violation. One in twenty-one carried an illegal weapon.
And that got the cops going: an arrest for fare beating wasn't just about writing a paper summons anymore. Now it was about making felony collars. And when fare beating went away, crime fell, and so, too, did the sense of disorder. And when it did, ridership returned. The MTA coffers began to fill again; the capital plan could go forward. That was the idea.
Take care of the small stuff, shake the tree for information, and you head off the big stuff. Take a fare beater or a low-level drug dealer off the street, and whatever criminal behavior he had in mind goes away with him. You can control behavior to such an extent that you can change it. That was the broken windows theory in practice.1
1 The "broken windows" theory was articulated by George L. Kelling and James Q. Wilson in the March 1982 issue of The Atlantic: "Broken Windows: The Police and Neighborhood Policing."
Les informations fournies dans la section « A propos du livre » peuvent faire référence à une autre édition de ce titre.
