CARVIEW |
Want a Map of Tehran? Use Open Street Map or Google
by Brady Forrest | @brady | comments: 0
All eyes are on Tehran right now. As the center of the Iranian election protests the city has become increasingly important to websites this week. To keep their site up-to-date with this latest crisis area Flickr switched out the Yahoo road Map with Open Street Map. When I heard about this I wondered how other major mapping sites faired.
So I examined the road and satellite maps of Yahoo, Mapquest, Google, and Bing (formerly Live Maps). Looking at the images below it becomes very clear that user-generated maps win in hard to reach places. Both Open Street Map (above) and Google (below) rely on user-contributions. Open Street Map relies almost entirely on user uploaded GPS tracks for its mapping data across the world. After the jump i've included the satellite maps from each service (except for Mapquest who did not have them). They were
Google is using data acquired from their just-under-a-year-old Mapmaker program (Radar post). With Mapmaker users can add roads, POIs, regions and features. It's a very powerful tool that has greatly expanded Google coverage. Google has been slow and deliberate in using Mapmaker data on their main site. In fact it was just a couple of weeks ago that Iran's mapmaker data "graduated" to the main site. There are now 64 countries on Google that have been updated with Mapmaker data.
This isn't the first time Flickr has done this (Radar post). They've also used Open Street Map for Beijing, Black Rock City (2008), Tokyo, Buenos Aires, Adelaide, Sydney, Brisbane, Canberra, Melbourne, Baghdad, Kabul, Kinshasa, Mogadishu, Harare, Nairobi, Accra, Cairo, and Algiers.
So what's holding back Microsoft, Yahoo and Mapquest? Unknown, but hopefully they'll realize that their top-down approach isn't working.
Compare the Maps for yourself:
Note: I have included data layers where they were available (Google and Microsoft).
(The markers include Wikipedia articles, photos, video, webcams, POIs, and public transit stations)
Bing Maps:
(The markers include Photosynths, user collections, photos and Wikipedia articles)
(This took a while to find, I had to find the International Maps page and click-thru a couple more pages to get the map)
tags: geo, google, osm, tehran
| comments: 0
submit:
Four short links: 17 June 2009
Word Mining, Open Ideas, Power Meter BotNet, and Realtime Web Traffic Tracking
by Nat Torkington | comments: 0
- NY Times Mines Its Data To Identify Words That Readers Find Abstruse -- the feature that lets you highlight a word on a NY Times web page and get more information about it is something that irritates me. I'm fascinated by the analysis of their data: boggling that sumptuary is less perplexing than solipsistic. Louche (#3 on the list) has been my favourite word for two years, by the way, since I heard Dylan Moran toss it out in that uniquely facile way the Irish have with words. I think Irish citizens get this incredible competence with the English language for free, along with staggering house prices and beer you can walk on.
- Open Ideas -- Alex Payne's blog of Concepts in the public domain, awaiting collaboration and appropriation.
- Buggy 'smart meters' open door to power-grid botnet (The Register) -- Paul Graham said that we've found what we get when we cross a television with a computer: a computer. Similarly, intelligent power meters are computers, computers that apparently haven't been well-secured. To prove his point, Davis and his IOActive colleagues designed a worm that self-propagates across a large number of one manufacturer's smart meter. Once infected, the device is under the control of the malware developers in much the way infected PCs are under the spell of bot herders. Attackers can then send instructions that cause its software to turn power on or off and reveal power usage or sensitive system configuration settings.
- Chartbeat -- the sexiest web analytics ever. It gives realtime count of users, whether they're reading or writing (based on whether focus is in a form element), where they're from, mentions on Twitter, and more and more and more. This is a different form of analytics than Google Analytics, which tells you trends and historical access. Love this for the pure sex appeal of a heads-up dashboard that can tell you exactly how many people are on your site and exactly what they're doing. (via Artur)
tags: analytics, crowdsourcing, data, energy, innovation, lazyweb, mining, security
| comments: 0
submit:
ARhrrrr! : Augmented Reality Zombie and Helicopter Game
by Brady Forrest | @brady | comments: 0
Augmented Reality is going to be coming to a phone near you very shortly. All it takes is a decent processor, a camera, a compass and a GPS -- all of which are becoming increasingly common on smart phones (Android phones and the iPhone 3GS qualify). It's going to be used primarily for games and geo-oriented apps. ARhrrr! is an awesome looking example of the former. In this game you hold your augmented reality enabled device above the "board". When you look at the board through your device you see buildings appear and ZOMBIES (left image).
In this game you are a helicopter (middle image) trying to save civilians. You can shoot zombies -- the zombies can also attack you with organs. Or you can use Skittles as bombs (orange and green plumes in the right image). The video below gives a great explanation of the game:
The game was created by Georgia Tech Augmented Environments Lab (also known for their virtual iPhone puppy) and the Savannah College of Art and Design (SCAD-Atlanta). As they describe it:
By merging graphics with props in the physical world, handheld Augmented Reality games pull the player through the small screen and into a larger merged play-space. Our primary motivation for this AR game was to explore fast-action first-person augmented reality, where the camera controls and movement that would typically require a mouse and keyboard are handled directly by simply moving the device. Advanced tracking technology allows the player to quickly zoom in and out and view the world at steep angles, making this a highly interactive and engaging game. Finally, we wanted to test tangible input mechanics, such as placing and shooting Skittles to trigger in-game events.
ARhrrr! t was designed for the Nvidia Tegra, which unfortunately doesn't seem to be widely available. I can't help but be reminded of Will Carter's Mobzombie game which uses AR to chase the player with virtual zombies. Luckily for all of us Mobzombies is coming to the iPhone (follow @mobzomies for progress) so you can get your virtual zombie fix.
tags:
| comments: 0
submit:
Interesting Questions Raised by Iranian Twitter Activism
by Timothy M. O'Brien | comments: 14
Development (4:10 PM CST): The State Department has been in contact with Twitter to make sure that the service remained available for protestors in Iran. (reuters)
Last Friday, Twitter started to digest the Iranian election results, and the tool became a powerful vehicle for protest and coordination for student protestors within Iran and interested parties outside the country. American Twitterers used the power of the medium to push our own media machine to increase coverage of the story via #CNNFail and #iranelection, and several dedicated observers did some important work to create proxies allowing the Iranian opposition to circumvent network restrictions. While it is amazing to see individuals using technologies such as Twitter to sidestep repressive government censorship, Twitter has also made it easier for observers, a world away, to become active participants in an unfamiliar political system at times taking vigilante action against the server infrastructure of a nation-state.
Figure: Graph of #iranelection from Twist.
tags: government, iran, protest, twitter
| comments: 14
submit:
Four short links: 16 June 2009
by Nat Torkington | comments: 5
- Dealing with Election Results Data -- taking the raw UK European election data into Google's Fusion Tables to try and make sense of it. More cloud-based tools for the data scientist within. (via Simon Willison)
- Time for an Open 311 API -- "311" is the US number to call for non-emergency municipal services. There have been a lot of individual projects to hack together web sites that provide the single coherent view of government services that the government itself is unable to offer, but the individual projects have all built their own APIs. SeeClickFix suggest these be unified so tools can be written (e.g., iPhone apps) that run across multiple municipalities. (via timoreilly on Twitter)
- Shoppers Cars Soon Able to Power Supermarkets (Daily Mail) -- At the Sainsbury's store in Gloucester, kinetic plates, which were embedded in the road yesterday, are pushed down every time a vehicle passes over them. A pumping action is then initiated through a series of hydraulic pipes that drive a generator. The plates are able to produce 30kw of green energy an hour - more than enough to power the store's checkouts. (via Freaklabs)
- Humans Prefer Cockiness to Expertise (New Scientist) -- the blogosphere explained in one paper. (via Mind Hacks)
tags: apis, brain, data, energy, google, gov2.0, visualization
| comments: 5
submit:
Walking the Censorship Tightrope with Google's Marissa Mayer
by James Turner | comments: 1
You may also download this file. Running time: 18:36
Subscribe to this podcast series via iTunes. Or, visit the O'Reilly Media area at iTunes to find other podcasts from O'Reilly.
Google sometimes finds itself at a difficult crossroad of wanting to make as much information available to as many people as possible, while still trying to obey the laws of the countries they operate in. I recently had a chance to talk to Marissa Mayer, who started at Google as their first female engineer, and has now risen to the ranks of vice president in charge of some of Google's most critical product areas, such as search, maps, and Chrome. We talked about some of Google's future product directions, and also about how Google makes the decision as to when information has to be withheld from the users. Marissa will be delivering a keynote address at the O'Reilly Velocity conference next week.
James Turner: As VP of Search Products and User Experience, you're responsible for a vast swath of the Google product line, from search to maps to Google Labs. You were also the first female engineer at Google. Can you talk a little about how you came to Google and what brought you to where you are today?
Marissa Mayer: Sure. My background is when I was at Stanford, I was doing a symbolic systems degree in artificial intelligence. And I was always somewhat interested in search. I ended up getting an email [from Google] towards the very end of my job interview process. I came to Google and did the interviews. And I came here because I really wanted to put my AI background to use. For about the first year or so, I did. I did a lot of work on categorization, some work on search quality. And then interestingly, we sort of had a void around how the site looked and felt and how it worked. And we tried very hard to hire someone in UI. We thought we needed someone to do UI like one day a week, and do systems engineering the rest of the time. After a few months of failing to hire such a person, Urs Hölzle, our VP of Engineering, pulled me in and said, "Marissa, we've looked through all of the resumes and you have this background in your undergrad on cognitive psychology and philosophy and things. And would you mind dedicating one day a week to UI?" So I did.
I pulled together a volunteer team to help out one day a week while we all still worked on our various AI and systems work. And then, of course, one day became two days which became three days or four days or five days. And I was also programming at that point. I switched over from a lot of the AI work I was doing to programming the front end for Google, working on the Google web server because it was nice for me to be able to not only make decisions about the UI but also to implement them.
And then because I was implementing the changes to the front end, I would go and meet with Larry and Sergey. And they would say, "What's happening on the site this week?" And I would say, "Well, I coded a change that looked like this. And I coded a change that looked like that. And translated this page and it'll go here. And there'll be a pull-down over there for the number of results." And without even realizing it, I did project management before Google had project management, by specifying how things were and looked and communicating to the rest of the company about how these changes would manifest. And then when we got to about 200 or 300 people, Larry and Sergey discovered that most companies have this function, product management, which I didn't know about; they didn't know about prior [to this]. And we decided we should have such a department. They realized there were a few of us around the company that were doing product management even though that wasn't our title. So they started the product management group and got it started earlier. And so I became a PM.
First, I was the PM on Google.com, really broadly across the whole site, because there were only three of us. One of us did the site, which was me. Salar Kamangar did the ads. And Susan [Wojcicki] did the partners. And then as our teams grew, I became the director of consumer web properties, where I still did all of the consumer facing work of the website including branching into Gmail and tool bars and some of these other areas. And then as we progressed, eventually, we restructured so we had search, ads, and apps, because Gmail and the related space of calendar and docs became large enough that it made sense to spin it out. So then I kept the search piece and the properties we have that are more related to search.
James Turner: There's always been other companies trying to take a piece of Google's dominant position in search; how does Google plan to stay a step ahead in search, especially in light of new players like Wolfram and Microsoft's new emphasis on Bing?
Marissa Mayer: Well, we are very focused on search. We have a large team here that's really focused on it and working hard on it. And we're constantly trying to forge in new directions. So we were really excited about the launch of our search options page because we think that allows us to try a lot of new ways to slice and dice and filter results. We were also really excited about Google Squared, which attempts to do automated text extraction from the web and present comparison tables for different entities in response to queries. They're both new ways to search. How do you generate a timeline from a web search? How do you generate a comparison table? And some of our competitors are also looking at those same issues. So I think on the whole, right now, our search is a very healthy ecosystem. There's a lot of interest. There's a lot of activity, and there's a lot of new ground being forged.
James Turner: Google users want the most useful results, but content providers want to get their pages seen, sometimes it seems at any cost. How will Google continue to provide the most useful results in the world of increasingly sophisticated SEO gaming?
Marissa Mayer: Well, we generally -- we really want to be fair in these issues as well as be good to users. We do think that spam is very detrimental to the user experience. We do have an incentive to find spam and remove it from our results. But we want to do something in a way that's very scalable. The web is scaling at an incredible rate. And we don't think it's really viable to try and fight spam in a manual way. So we're always looking for new algorithmic ways to understand new spam techniques, to be able to detect them in an automated way and remove them from our results. And the nice side benefit that scalability has is it's also reasonably objective and fair.
tags: censorship, google, interviews, maps, news, seo
| comments: 1
submit:
Personal Democracy Forum ramp-up: from vulnerability and overload to rage, mistrust, and fear
by Andy Oram | @praxagora | comments: 3
The Personal Democracy Forum will hold its sixth annual conference at the end of this month. The theme, "how technology and the Internet are changing politics, democracy, and society," has been central to O'Reilly's work over the past few years (and a theme on which we're holding a summit of our own in September). Over the next two weeks I'll write three blogs on the Radar site to get some of my current thoughts off of my chest, clearing some space so that when I get to the PDF conference itself, my blogs can focus on its events and statements made by its participants.
This blog covers:
- The government participation cycle: if you want to dance, sir
- Vulnerability: a reason to put brakes on outgoing information
- Overload: a reason to put brakes on incoming information
- Rage, mistrust, and fear: inhibitors of the government participation cycle
The government participation cycle: if you want to dance, sir
The grand vision for government/public collaboration is a set of feedback loops that intensify the influence of the collective will on government policy. A feedback loop might consist of a cycle like this:
- An agency (or less likely, a legislature) posts data in a downloadable format through a flexible API and announces a call for applications.
- Companies and public interest groups define goals and put programmers on the task.
- The public uses the resulting applications to generate data and share it with the agency.
- The agency sets policy or changes direction in response to the data.
At any step, a failure by any of the responsible actors to follow through will leave the process hanging and discourage future projects in open participation.
This doesn't mean every project needs to include all four steps. The public may benefit from government data without offering feedback, and programmers could put their work under an open source license or into the public domain for the benefit of the government or members of the public without asking them to share more data. Agencies can also use programs to improve internal coordination instead of working with the public. But the full four steps serve as a canonical model for government/public collaboration.
Successful examples already exist for each step. As I write this, Data.gov has 261 data sets and 30 tools; thousands more data sets are promised soon. Appeals for donations of code, such as Vivek Kundra's Apps for Democracy in Washington, DC and the Sunlight Foundation's Apps for America, show that coders will play their part, at least in the current atmosphere of enthusiasm for the new initiative. And the public has responded to requests for data.
But at the federal level, we need to dance a few rounds of the full cycle before feeling confident that open processes are fully entrenched. I'll return to this theme in the last section of this article. The cycles of public participation will teach lessons, of course, that feed into a still larger cycle of constant experimentation and improvement.
As public participation moves forward, it's worth remembering that resistance to the free flow of incoming and outgoing information is not irrational. The resistance spring from healthy coping mechanisms learned by individuals and organizations learned over their lifetimes. I have already published and solicited comments on a list of fundamental questions on government participation; in this article I describe two such issues that play a special role in resistance to information sharing.
Vulnerability: a reason to put brakes on outgoing information
A couple months ago, I read a stirring report from a federal agency manager trying to sound out the Administration concerning how much the agency ought to reveal. The manager was stunned and inspired by the response of Bev Godwin, a prominent director at the White House and General Services Administration, who advised talking about the bad things as well as the good and soliciting negative as well as positive feedback.
Vulnerability is the keystone of transparency and openness. Online forums, if they are run democratically and competently, encourage vulnerability through a combination of self-correcting mechanisms:
- The right to respond
- Anyone criticized in a forum has repeated chances to defend himself at length. If the forum includes a rating system, persuasive arguments and well-chosen facts will float above false accusations as well as flaccid excuses from the accused.
- Support networks
- Proponents of each side pile on to each debate, turning it into a community issue and diluting the personal biases brought by the people who began the debate. A bit of a mob scene can erupt at times, awakening the risk that the losing side will walk away in a huff while sensitive community members flee the fury. But as long as participants value the community over partisan agendas and prefer honesty to grandstanding, the community comes out stronger, more aware of its options, and ready to integrate what it has learned into further action.
- Community memory
- Forum members recognize when old debates are re-ignited, and can fill new members in on the history. They can also predict the way prominent participants will line up on an issue. Debates are thus tighter and more quickly resolved.
- A propensity for truth
- These traits all end up privileging accuracy and making it harder (although not impossible) for bad judgment to prevail through false claims, manipulative demagogy, appeals to group solidarity, and the other tricks used by insincere factionalists.
This list may present online forums in a bit too rosy a light. But they do permit social norms that protect vulnerable people, even if the norms don't function perfectly. The real problem comes when words leave these forums and end up in other environments not subject to the same rules.
Government staff have already witnessed too many negative experiences in traditional, non-virtual settings. They have seen what happens when a comment is taken out of context and bandied about in the broadcast or print media, introduced into court testimony, or used as ammunition in partisan debates. They know that comment posted on the Web can be fodder for the same opinion machine--and are in fact even more dangerous because the Web makes them more visible.
That's not fair. It's very hard for anyone outside an agency to judge why it came down on one side of a debate or what that decision's long-term effects will be. Most agency actions are a complex fermentation blending the data that was gathered, assessments of the data's accuracy, assessments of the possible trends indicated by the data, consultations with the public (yes, outsiders are routinely consulted), judgments about Congress's intent, judgments about the interests of the Administration, and more. But groups with a cause like to ascribe one-dimensional reasons for key agency decisions and mine public statements for corroborating evidence.
This doesn't mean that all agencies are honest and act in the public interest. Plenty of bad government decisions have been made under pressure from well-organized special interests or to pay off political donors. One role of civil society is to expose these influences--that's what open government and the Personal Democracy Forum are all about. So we want more of these online forums. But we also need to protect the agencies whom we expect to use the forums. To encourage the necessary vulnerability, we have to combat those who abuse the results.
Journalism is starting to incorporate its own feedback loops and open its pages. Elections and policy debates are also monitored by the blogosphere. So some forums are becoming friendlier to the cause of vulnerability (the court system is unlikely ever to change). But it will be a long time before it's safe to lay out one's thoughts in an open, self-policing community.
Overload: a reason to put brakes on incoming information
The previous section mentioned the possibility of a "mob scene," and if people putting out information must be able to tolerate being vulnerable, those requesting input from the public have to deal with a potentially low signal-to-noise ratio.
We need not look far for an example. Take last month's brainstorming session on open government, launched by the White House and the Office of Science and Technology. It drew over 1,000 submissions in a single week. (Even more are on the site now, but they arrived after the official close of the session.)
The thousand submissions offered quite a smorgasbord for a group led by the new Deputy Chief Technology Officer for Open Government, Beth Noveck, to spoon through. They ended up with many intriguing ideas. But the gathering of ideas was simply a suggestion box, not real crowdsourcing. The web site offered no tools for editing, combining, and culling entries (and there would be inadequate time to use such tools anyway). The only aspect reminiscent of group behavior was a casual and anonymous rating system, which played little role in the results.
And that's a relief. After all, how many Americans would be able to assess the Office of Open Government created by Florida Governor Charlie Crist, or the potential for Cooperative Research and Development Agreements to help convert government data and applications to open source? Both of these projects earned a place in the results, even though the Florida model got only 24 votes and the Cooperative Research and Development Agreements only 46. (Although they might have conceivably been mentioned in an earlier brainstorming session conducted among government workers, I couldn't find them in the publicly posted comments.)
In response to a question about the voting, Noveck wrote me, "We wanted to encourage the National Academy of Public Administration to try different voting techniques. They started out by allowing voting by unregistered users, and later restricted it to registered users. Given the change, we didn't want to disadvantage anyone who participated. Consequently, we viewed the voting as informative but not determinative. On our weblog, only registered users can vote on comments."
As her statement indicates, the second phase of this transparency project has already sprouted more of the checks and balances found in mature discussion forums. We can expect the Administration to wend its way toward systems that gather useful opinions from self-organized groups of qualified commentators, the model pioneered by Noveck in her Peer to Patent project.
But will the White House have the time and resources to establish a foothold for a solid and lasting open government program? That depends on public tolerance for the Administration as a whole.
Rage, mistrust, and fear: inhibitors of the government participation cycle
Everyone knows that productive collaboration can't take place under conditions of rage, mistrust, or fear. Americans unfortunately are suffering from all these feelings right now.Their rage has been directed at the heads of the financial industry. No peasant at the time of the French Revolution felt more hatred for Marie Antoinette than some of the comments I've seen about AIG. In addition, the current conditions of recession and financial uncertainty breed mistrust toward all three branches of government, and fear toward anyone who could seem to wangle an extra advantage over other Americans.
I'm not going to factor in the recent murders of law enforcement officers, Dr. George Tiller, and others because I'm sure the hate crimes were caused by lots of diverse factors, and it's unclear whether they represent a widespread cultural movement. We have plenty to worry about just by considering problems that will undeniably have a broad impact on Americans.
Over the coming year, lots of homes will continue to be foreclosed (because Congress failed to put a system in place to stop them), a blight that hits many neighborhoods like a dry Katrina. This ongoing crisis will be joined by credit card crisis (because Congress's bill didn't do much to stop that either) and perhaps already a student debt crisis. The Administration has its own challenges, waging two untraditional wars that nobody knows how to win and tinkering with a global financial system that always cracks its casings.
Open government doesn't deserve to be at the mercy of current political controversies. It did not originate with the Obama administration, and it doesn't require a Democratic Party philosophy. The George W. Bush administration took some steps toward open government (often forgotten amongst all the complaints over their unsavory maneuvers and information withholding). The Bill Clinton administration took steps too. But Obama is making it a centerpiece.
This gives us more hope than ever for openness, but ties its fortunes to the larger sphere of activities by the Administration and federal government.
To establish a foothold, openness needs some early, impressive success stories. Federal CTO Vivek Kundra has said his initiatives will prove themselves by saving money, although that certainly isn't his sole aspiration. If the Administration can land a few universally recognized successes--budgetary or otherwise--and especially if it can run through the whole cycle I laid out at the beginning of this article, such efforts will be continued by future Administrations.
Next article (Friday, June 19): twenty-five hundred years of Government 2.0.
tags: democracy, governance, government 2.0, open government, personal democracy forum, transparency
| comments: 3
submit:
Hands-On with the iPhone 3.0 OS; Search is the Winner
by Brady Forrest | @brady | comments: 4
The iPhone 3.0 OS is going to be released this Wednesday. It will be available to all iPhones (for free) and iPod Touches (for a small cost). The iPhone 3GS will ship with it. The new OS became available last week to those willing and able to try it out a bit early (see Gizmodo for details). This is what I've noticed since updating:
Search - The new OS has search everywhere. It's the zero screen (to the left of your homescreen), in mail and in the iPod. It has become my preferred way of launching applications and finding contacts. You can choose what you want to search and their priority in Settings->General->Home (screenshot). I immediately set the Home button to go to Search, so that with just two taps I am able to launch an app. Unfortunately, there is no spatial component to contact search (I'd like to be able to pull up everyone in a town I am visiting for instance. I hope that third-party apps will be able to include their content in Search.
Voice Memos - Voice recorders have been around since the App Store launch so I didn't really understand why Apple was spending development resources on their own version. Now I know, Apple's version will run in the background letting you access other apps while getting an audio record of your surroundings. Unsurprisingly, you can't record phone calls with it.
Cut-n-Paste - The iPhone finally includes cut-n-paste. Initially I found it awkward, but now that I've figured out that double-tapping any text to select it. It is really, really nice to have this feature -- almost as nice as getting Search.
New iPod Controls - The iPod will now let you jump back 30 seconds if you missed something. It will also let you speed up or slow down audio. This is very nice for us appreciators of the spoken word.
Safari - Unfortunately I have not found any web app that uses the new geolocation and offline features of the new Safari browser. I expect that apps like Google's Latitude that use geolocation (demoed working in the browser at Where 2.0) and the new GMail that uses offline storage (demoed at the Web 2.0 Expo). I have not perceived the new Browser to be appreciably faster
There are other features (like forwarding SMSs and more detailed call logs), but these were the most significant updates to me. Right now there is no way the new Hardware API or P2P API. We'll have to wait for new apps to come to the store to try them out. I'll be getting the new iPhone 3GS this week and will share my impressions of the new hardware shortly.
tags: mobile
| comments: 4
submit:
Jeff Bezos at Wired Disruptive by Design conference
by Tim O'Reilly | @timoreilly | comments: 18
Jeff Bezos is very quotable. Listeing to Steve Levy interview him at the Wired Disruptive by Design event in New York, I was furiously taking notes. Here are the quotes I managed to capture:
"We've co-evolved with our tools for thousands of years," he says, explaining how ease of Kindle buying changes behavior.
"Reading is an important enough activity that it deserves a purpose-built device....It's a myth that multi-purpose devices are always better.... I like my phone... I like my swiss army knife too, but I'm also happy to have a set of steak knives."
"I get grumpy now when I have to read a physical book....The physical book has had a great 500 year run, but it's time to change."
"If you're an incumbent in any industry, and rapid change is underway, you're uncomfortable, even if long term it's going to be good."
"I want to be able to provide every book ever published, in any language, in sixty seconds." Steve Levy then asks him about the GBS settlement, but Jeff declines to comment. But adds nonetheless: "That settlement needs to be revisited, and is being revisited....It doesn't seem right that you can get a prize for violating a large set of copyrights."
Steve Levy: Going back to disruptive internet companies from 1990s - "could an established company have understood how to build a great internet company?" Bezos: "One of the statistics I saw in 1994 that encouraged me to start Amazon was that net usage was growing 2300% a year. But it was still tiny....One of the biggest problems with big companies doing clean sheet innovation is that even if you see it, you have to be a really long term thinker, because for a long time it will be a tiny slice of the company.... The key thing is to be willing to wait 5, 7, 10 years. And most companies aren't willing to wait ten years."
Steve asks him about how he survived the years in which everyone doubted his strategy. Jeff replied: "There were two things: the business metrics, and the stock price. After the bust, the stock price went down, but the business metrics continued to improve....We had some very harsh critics during that time, but we always noticed that our harshest critics were among our best customers. Having a team that is heads down focused on building product makes you more resilient against outside opinion."
During the 1999 stock bubble, Jeff says he kept telling employees "Don't feel 30% smarter because the stock is up 30% this month, because you'll have to feel 30% dumber when it goes down."
"One of the differences between founder/entrepreneurs and financial managers is that founder/entrepreneurs are stubborn about the vision of the business, and keep working the details. The trick to being an entrepreneur is to know when to be stubborn and when to be flexible. The trick for me is to be stubborn about the big things."
He also talks about basing business strategy on things that aren't going to change: "I know that ten years from now, customers are still going to want low prices. I know they are going to want fast delivery. I know they are going to want the biggest product selection."
"There are a few prerequisites to inventing.... You have to be willing to fail. You have to be willing to think lng term. You have to be willing to be misunderstood for long periods of time. If you can't do those three things, you need to limit yourself to sustaining innovation.... You typically don't get misunderstood for sustaining innovation."
"At the end of the day, you don't end your strategy because other people don't understand it. Not if you have conviction."
"These new businesses are very energizing. We don't 'stick to the knitting'...I wouldn't even know how to respond if someone said 'Jeff, this isn't the knitting.' But we do make business decisions in a very deliberate way: we work backwards from customer needs, and we work forwards from our business skills."
"You've got to be willing to learn new skills if your customers need you to have those skills."
"18 months ago we launched Kindle with 90,000 titles. It's 300,000 now, and it's accelerating."
"We've made many errors. People over-focus on errors of commission. Companies over-emphasize how expensive failure's going to be. Failure's not that expensive....The big cost that most companies incur are much harder to notice, and those are errors of Omission."
tags:
| comments: 18
submit:
The Four Pillars of an Open Civic System
by John Geraci | comments: 9
Everyone is talking a lot about open government and transparency these days. It's exhilarating stuff, and it's even more exciting to see governments get behind it, creating sites like data.gov in the U.S. for the public to access government information via APIs. But every time I hear someone say something like "our organization is really into transparency" (which is often) it sounds odd to me. It's only talking about a part of the puzzle, not the whole puzzle. What we really want (or what I really want anyway) is not simply government transparency, but an open civic system - a civic system that operates, and flourishes, as a fully open system, for whatever level we happen to be talking about - federal, state, city, neighborhood, whatever. And transparency is a big part of that open civic system, but it is still only one part.
In fact there are four parts to a functioning open civic system. These are:
Government to Citizen (G2C). This is what people speak of when they talk about transparency and open government data. It's the idea of creating open pipelines for data directly from government and gov't agencies to whoever is interested in receiving it. G2C gets you accountability - watchdog groups suddenly have easy access to the paper trails for everything that is going on, etc. It also gets you things like transit schedules, minutes from meetings, and zoning data - things that can be built on by third parties to make the civic system work better. G2C is critical stuff, but without the other three components in place, we can't make the most of this open government data. What we need is not simply a pipe of open data, we need an ecosystem of open civic data, all interconnected, all flowing every which way. That's what the other three "pillars" of an open civic system gets us.
Citizen to Government (C2G). The counterpart to G2C. This is the idea of creating open pipelines from the people directly to the government - hopefully with someone listening on the other end. Adding C2G to G2C completes the circuit and makes open government APIs and such that much stronger - it takes what was a uni-directional data flow and turns it into a feedback loop of information, input and output. At the city level, C2G is taking shape right now in the form of Open 311 - a open API that anyone can build on that allows residents to create "problem tickets" for their city to address one way or another. Washington D.C. is currently launching an open 311 API, and I expect more cities will follow suit soon. Other examples of C2G include UK's FixMyStreet and SeeClickFix.com from New Haven, Connecticut, both sites with a huge amount of potential. There are a million different, nuanced ways C2G could be played out, at the local, state and federal levels.
Citizen to Citizen (C2C). Okay so now we have both open G2C and C2G data flows going, and that's great - huge amplification of civic activity, great realization of efficiency with regards to interaction between government and people. But there are all sorts of ways to improve civic life that don't really need to involve the government at all - what about those things? That's where Citizen to Citizen, or C2C, data flows come in. C2C is the citizens' brigade of data flow - it's the people doing it for themselves, whatever "it" happens to be. Clever Commute, in New Jersey, is one example of a great C2C data flow. Everyone who commutes by train into NYC subscribes to the Clever Commute feed, and then notifies each other of what the current delays are, and where, each morning. The system works better than anything New Jersey Transit has been able to pull together, and at a cost of essentially zero. This is the great thing about C2C - it is added value to the civic system at no additional cost to the system itself. The cost to operate C2C is passed on to those who are using it, and spread out amongst individuals, to the point where the costs become negligible. Instead of New Jersey Transit coming up with a system that knows how late each of its trains are at a cost of millions of tax dollars, the users of Clever Commute bear the cost of the system, and it costs pennies for each user to operate (the cost of sending a text message). C2C is a huge value-add on top of G2C and C2G, and as governments consider how to get increased services in these recessionary times, I expect C2C to be huge - once governments get used to the idea.
Government to Government (G2G). Lastly, the square is not complete without open Government-to-Government data flows. Entities within governments should have easy, open data exchange with each other, without having to issue a request, parse something out of a PDF, and so forth. The ability for, say, the NYC Department of Health to get data from the Los Angeles DoH in realtime, without having to talk to anyone or issue a request could be a huge asset. Or think of the efficiencies that could be gained if the NYC DOT were able to exchange realtime data with the NYPD. If these examples sound vague, it's because G2G is the "pillar" I know the least about, having never worked in a government agency. From what I've learned though, it seems to me that there could be a huge increase to civic utility with a little bit of thought about an open G2G system.
And of course you can blend these data flows and come up with hybrids all you like. DIYcity's SickCity, for example, is basically a C2C tool in its present, basic 1.0 incarnation - it detects instances of residents in your city saying they're sick, and passes that news on to other residents. But a more sophisticated version of the tool would also pass that information on directly to the Department of Health when relevant, and would also, optimally, accept data from the DoH to pass that back to residents. Suddenly it has gone from a simple C2C tool to a tool that is C2C, C2G and G2C. Now we're talking about interesting stuff. Each additional channel of data makes the system exponentially more valuable.
With all of these systems properly developed and engaged, our civic systems - local, regional, federal - should bloom and transform into the properly modern, Internet-age things they ought to be. This will translate to increases in efficiency, greater innovation and rate of change, better adaptability, and greater resilience, in addition to other advantages. To get there though, we've got to get beyond thinking simply in terms of transparency and government APIs.
tags: open civic, open data, open government
| comments: 9
submit:
Four short links: 15 June 2009
by Nat Torkington | comments: 0
- More Talk Less Chalk -- wordy slides that duplicate what the speaker says make it harder to learn. [R]esults indicate that participants exposed to lexically sparse slides had better recall of thematic content, suggesting that deeper encoding occurs when working memory demands are reduced, and that this may be achieved simply by minimising the number of words on the slide. In presenting competing visual and verbal lexical narratives, instructors may be inadvertently increasing the extrinsic cognitive load. If you don't think that your job requires you to know how to communicate, you're wrong. (via Titine on Delicious)
- AP in Deal to Deliver Non-Profits' Journalism (NYTimes) -- Starting on July 1, the A.P. will deliver work by the Center for Public Integrity, the Investigative Reporting Workshop at American University, the Center for Investigative Reporting, and ProPublica to the 1,500 American newspapers that are A.P. members, which will be free to publish the material. A six-month experiment, the four groups combined have more than 50 professional journalists. Good to see media companies experimenting with new models for funding journalism. (via timoreilly on Twitter)
- Number of iPhone Subscribers is 6.4 Million -- Matt Gross reports on Nielson's iPhone numbers. Forty-percent of iPhone subscribers have household incomes of $100,000 or more, compared with 19% for all subscribers. 75% of iPhone subscribers download apps.
- Adam Savage: Colossal Failures -- video of his talk from Maker Faire, telling juicy stories" about his failures. Very watchable/listenable. (via waxy)
tags:
| comments: 0
submit:
XKCD on the Future Self
by Brady Forrest | @brady | comments: 12
This morning's XKCD, Latitude, spells out one of the reasons people will be weary of setting up continuous location trackers: the future self. The future self forgets that they are sharing their location and then act as if no one knows where they are going. In this case Megan's friend tracks her stops at a sex shop, toy store, hardware store and finally the burn ward, telling a pretty clear story of a mistranslated kama sutra (ahem).
The name Latitude comes from the recent Google service that will share your location with your network (or publicly). However, it could have just as easily been called Loopt, Brightkite, Fire Eagle or any number of other location-updating services. I am personally looking forward to these services becoming ubiquitous so that I can track my location and aspects of my life, but as they currently stand these services are not poised for mainstream adoption. They need to do more to prevent people from embarrassing or endangering themselves.
One location-sharing service that takes a more constrained role is Glympse. It allows you to share your location on an ad hoc basis with specific people for a specific amount of time. Check-in services like FourSquare (or Dodgeball) let you specify when you self-locate as you wish. These are half-measures and don't meet everyone's needs, but they provide important steps in the right direction.
tags: geo, xkcd
| comments: 12
submit:
Recent Posts
- Four short links: 12 June 2009 | by Nat Torkington on June 12, 2009
- Mechanical Turk Best Practices | by Ben Lorica on June 11, 2009
- Search for Developers | by Brady Forrest on June 11, 2009
- Four short links: 11 June 2009 | by Nat Torkington on June 11, 2009
- Programming Contests, Community, and Business | by Simon St. Laurent on June 11, 2009
- The King is Dead, Long Live the King | by Mike Loukides on June 10, 2009
- Clarke and the Continuous Location Update | by Brady Forrest on June 10, 2009
- John Viega Talks About Beautiful Security | by James Turner on June 10, 2009
- Four short links: 10 June 2009 | by Nat Torkington on June 10, 2009
- Twitter is Not a Conversational Platform | by Mark Drapeau on June 9, 2009
- Twitter Approval Matrix | by Mike Hendrickson on June 9, 2009
- Four short links: 9 June 2009 | by Nat Torkington on June 9, 2009
STAY CONNECTED
TIM'S TWITTER UPDATES
CURRENT CONFERENCES

Velocity brings together representatives from industry leaders like Google, Facebook, Microsoft, eBay, AOL, MySpace, and more — people who are doing the best performance and operations work in the world — to improve the experience of web users everywhere. Read more

Join us at OSCON 2009, the crossroads of all things open source, a deeply technical conference that brings community ideals face-to-face with business practicality. Come together with 3,000 of the best, brightest, and most interesting people to explore what's new and to help define, maintain, and extend the identity of what it means to be open source. Read more
O'Reilly Home | Privacy Policy ©2005-2009, O'Reilly Media, Inc. | (707) 827-7000 / (800) 998-9938
Website:
| Customer Service:
| Book issues:
All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners.