• the number of insurance claims on Homeowners, Commercial, and Auto policies,
• the number of paid and denied claims,
• how long it took the insurance companies to respond,
• the number of adjusters in the field, and
• the number of people filing complaints about their experience with the system.

NY State is constantly updating the spreadsheet:  http://www.nyinsure.ny.gov/ReportCardAllCompanies12-20.xls (last updated Dec 20), so I took a look at the data to see what’s going on.

Some key numbers:
372,435 total home insurance claims 
New York State accounts for about one third of the total home insurance claims for Sandy with another one third from New Jersey (although New York accounts for 51% of the total auto losses!). In total, New Yorkers claimed more than $2 billion of home losses — and at least the same again in un-insured losses . The total damage for Sandy has been estimated at $52 billion, at least half of which is not covered by insurance.

23.3% claims closed without payment
Almost 1 in 4 claims from Hurricane Sandy were deemed to be not worth paying. We don’t know if these claims were minor, not covered or even fraudulent — but this number highlights how important it is to have human oversight that checks every claim.

11 days average wait for claim inspection
Every day a person spends waiting for an inspection adds another day until they receive payment and can start rebuilding their life. Although more than 5,000 claims adjusters were working in the region, it takes time to get to everyone.

10,952 years total time spent waiting for inspections
Adding up all these delays, over all the people claiming insurance, victims of Hurricane Sandy waited a huge amount of time for inspections. With the amount of suffering in the region, insurers should be aiming to reduce this number to make the process more efficient and painless for the customer.

These data show the huge impact that Hurricane Sandy had on property, finances, time and lives. One thing is for sure: the ability handle post-disaster claims faster and more accurately means big wins for

• saving time,
• saving money,
• helping victims and satisfying customers,
• getting a passing grade on New York’s (and other states) public report cards.

With rapid access to satellite and aerial images in the aftermath of a disaster and crowds of people willing to help identify damage, it is now possible to get a handle on damage much faster than before. Next time disaster strikes, satellite image information can be used to accurately and quickly process a large number of claims. This would provide a head-start to insurance adjusters and a better service for consumers.

If it meant a faster payout, would you let your home damage be assessed from space?

Overhead image showing damage in Rockaway NY

I spoke about how Tomnod “crowdsources the world” by combining real-time satellite images of the planet with large-scale human perception. But I wanted to add a personal touch… so I tied in my love for travel, video games, sneakers and even a wedding picture of my beautiful wife Laia!

Watch the talk here:

Speaking in front of 600 people in the packed City Performance Center in Dallas’ Arts District was a nerve-wracking thrill. But just as exciting were the other fascinating, educational and inspiring speakers. Louie Schwartzberg films the mystery of the earth’s tiniest creatures and David Gallo took us to the deepest trenches of the ocean. Jeremi Suri taught us how to use history to understand our future while Christian Genco convinced everyone to learn to program! The day was filled with memorable music too: Katrina Leshan trasnsformed visual art into musical art and Lewis Warren interpreted Star Wars, as composed by Chopin!

Our hosts on stage brought a touch of TED class to proceedings: as talent coordinator for the main TED, Kelly Stoetzel knows how to set the stage for a great talk. Her co-host Rives is a prodigy of the presentation — you’ve got to check out his TED talks on Wikipedia wanderings and the 4am conspiracy!

The quality on stage was matched by the backstage organization. Sharon Lyle and Heather put together an amazing show. I also loved how each talk was “live-scribed” by CollectiveNext: as I spoke, my words were translated into the amazing cartoons reproduced below.

My sincere thanks to the Young family, especially Jim and Laura, for their great support and hospitality and for organizing a world-class TEDx event. And, as always, I couldn’t have done it without the support of my fellow silver-sneaker-wearer and beloved wife, Laia Egea Pujol!

If you have any comments about my talk — or my silver sneakers, post them below!

Satellite view of Gaylord Palms

Last year was Tomnod’s first GEOINT when we were invited by USGIF to receive the 2011 Research Achievement Award for our work in leveraging crowdsourcing and remote sensing to find the tomb of Genghis Khan. Realizing that this event was not be missed, we made sure that this year was even better.

Early on in the week we teamed up with DigitalGlobe to show off our new FirstLook+ capability, where we add a layer of crowd-derived insight onto DigitalGlobe FirstLook imagery. We are really excited about helping emergency management personnel, government officials, and effected industries be more efficient in leveraging satellite imagery as they respond to crisis events around the world. Combined with our presence on the main stage during the Young Innovators Panel, we had a great show!

With that here are my top 4 reasons why GEOINT 2012 Rocked My Face Off (RMFO!):

1. USGIF – The entire conference was incredibly well put together and USGIF is a well-oiled machine. From GEOINT, to industry days, to the show daily magazines, USGIF is probably one of the strongest industry organizations I know (even compared to the massive IEEE!). I was also particularly impressed by how the Young Professional Group (YPG) in USGIF was given two board seats to make sure that the next generation of innovation has a voice within the organization.

I should mention how impressive YPG was. We teamed up with YPG to display a TweetMap at the YPG lounge. We tracked and mapped all the GEOINT conversation on Twitter. Tomnod’s GeoAPI geolocated every #GEOINT tweet, based on either embedded GPS or the name of the company mentioned. For each tweet-location, we pulled and chipped a recent satellite image of that location (from DigitalGlobe satellites). Social media directs satellite imaging for real-time geoint!
Congratulations to the winners of our tweetmap competition: 1) GeoEye 2) DigitalGlobe 3) Wiser

Our own Dr. Nate Ricklin (aka “The Rock”) standing proud in front of the YPG display

2. Coopetition – Throughout the symposium, there was a feeling of cooperation just as much as there was competition between the various companies. The sense was that companies were more interested in advancing the field and supporting the warfighter than just beating each other of existing market share. Don’t get me wrong, there was definitely competition however it was more based around “how can we make ourselves better” rather than “how do we make our competition worse”. Very healthy.

3. Tight Community – Everyone seems to know everyone at GEOINT regardless of which company they currently work for. Perhaps it is just the small number of people that are involved in the industry but there was a definite feeling of camaraderie between people at the various companies both at the conference and at the social events in the evenings. Good fun!

4. Spirit of innovation – I was invited to participate in the young innovators panel along with some amazing people who are at the top of their game. On the panel we talked about some of the challenges we face trying to innovate within the current system as well as the opportunities we see. We had a great discussion and I had a lot of lively follow on conversations with government officials and executives at large companies who are embracing that spirit of innovation. Rather than lamenting potential reductions in government budgets, people on the exhibit floor were looking forward to changing the way they operate and creating innovative ways to cut costs and improve performance. I am very bullish on the industry and the capabilities it is going to embrace and create in the coming years.

Young Innovators Panel. Image courtesy of @geointer

In conclusion: great people, great technology, great event. Over the next year I challenge you to embrace that spirit of innovation within your organization and within yourself. Innovators need just the right blend of challenge and support to succeed.

Let us be hungry but don’t let us starve!

Here is a video of the young innovators panel:


See you in Tampa!

Alas, the search ended in tragedy. Gil and Ben’s bodies were found on July 28, 2012.
But we learned a great deal about how to use Tomnod technology to support future search and rescue scenarios. Our vision is to institutionalize the process and streamline for rapid deployment in situations where a few hours can be the difference between life and death. To map out exactly what happened and learn how to do better next time, we present the following timeline of events.

KEY TIMELINES

• 30 hours to task, acquire and access current satellite imagery
• 2.5 hours to process imagery and display online
• 30 minutes to get the Tomnod crowdsourcing platform up and running
• 1 hour to find the first object of interest in the image
• 8 hours to sign up 288 volunteer taggers

GROUPS INVOLVED

• Friends & family (Purple)
• Ground crew in Peru (Red)
• Satellite sources (Orange)
• Crowd (Green)
• Tomnod (Blue)

Weds, July 25
0h	06:45	Email circulates from a climber friend about 2 missing climbers.
	Gil’s flight leaves tomorrow from Lima. He should be down from the mountain by now.
4h	11:00	Maps, online info and ground sources verify the GPS location and confirm that the crisis is real.
5h	12:15	Tomnod contacts partners at DigitalGlobe and GeoEye to ask for archival imagery of the mountain and request a new satellite collect to search for the climbers’ current location.
	12:30	DigitalGlobe and GeoEye respond rapidly, generously agree to make imagery available and start searching archives.
7h	13:54	Contact with US embassy in Peru.
9h	16:00	Ground search team plans to search in the early morning.
Thurs, July 26
24h	06:30	Positioning of commercial satellites scheduled. New data may not be available for up to 3 days.
26h	08:00	Ground search team gets to base of route; finds basecamp but no evidence of the climbers.
	After several hours of hiking, we found their camp near the West end of the glacier. Their things were strewn around almost exactly like they had left for a climb and simply never come back. We spent the rest of the day walking around on the glacier, following tracks and looking up various potential routes with binoculars. At this point we had almost nothing to go off of except some 2 week old tracks. Mid afternoon, I got information that they got into trouble on the decent, via a ridge to the West of us. We hiked up into a col to look at the ridge. Immediately, we saw tracks leading from the summit and disappearing halfway down. The end of the day was drawing near, so we descended to Ben and Gil’s camp, packed up their things and descended to the valley floor.
30h	12:16	Archival satellite imagery taken on July 2 (1 week before climbers depart) is sent to Tomnod.
31h	13:51	July 2 image is tiled and published to the web.
34h	15:43	As the story spreads, the request priority for current satellite imagery increases.
Fri, July 27
50h	08:00	Confirmation that DigitalGlobe satellite acquired images of the mountain from July 26th.
	08:14	Tomnod receives a preview of unprocessed satellite images from the 26th.
52h	10:30	Activation. To examine the large area covered by imagery, Tomnod decides to deploy crowdsourcing portal.
	10:45	Engage back-end platform to collect crowdsourced info: specify features to search for (person, tracks, camp, avalanche), define geographic area, start building the crowd.
54h	12:31	Tomnod receives 2.5Gb of raw satellite imagery from the 26th. Preview for imagery from the 27th is also sent but too cloudy to use.
57h	15:43	Processing and tiling complete and imagery made available in Tomnod portal.
58h	16:00	Crowdsourcing portal is launched at http://tomnod.com/gi/peru/
	16:10	Share link to crowdsourcing portal on Tomnod Twitter and Facebook pages, SuperTopo climbing forum, CrisisMappers email list and personal emails.
59h	17:00	36 registered taggers identified 329 image locations.
	17:39	3-man search and rescue team spotted in satellite image.
		3-man rescue team on the snow.
60h	18:00	62 registered taggers identified 570 image locations.
59h	19:00	89 taggers identified 871 image locations.
	19:49	Climbers’ basecamp spotted in satellite image.
		Climbers’ basecamp: left satellite image, right ground photo.
62h	20:00	114 registered taggers identified 1,245 image locations.
	20:15	Tomnod analysts use crowdsourced clues, before/after imagery and mountaineering expertise to prioritize top 3 likely locations of the climbers.
64h	22:00	169 registered taggers identified 2,538 image locations.
	22:30	Tomnod call on Facebook page to translate analysis results into Spanish.
	22:47	73 volunteers translate the report into Spanish.
65h	23:00	186 registered taggers identified 2,879 image locations.
	23:30	PDF report detailing locations and images of priority search locations sent to ground team in Peru.
		Tomnod report sent to rescue team, detailing locations and tracks identified from crowdsourcing.
Sat, July 28
69h	03:00	Rescue team leaves basecamp to begin exploring ridge under clear, starry, wind-free night.
	03:00	222 registered taggers identified 3,447 image locations.
72h	06:00	Aerial crew reviews Tomnod report, plans search mission.
	06:00	236 registered taggers identified 3,633 image locations.
74h	08:00	Plane leaves Lima.
	08:00	236 registered taggers identified 4,168 image locations.
75h	09:00	Start running Tomnod CrowdRank to identify consensus locations in crowdsourced data collected overnight.
76h	10:00	Email from Ben’s dad:
	Just heard that both Ben and Gil were spotted dead in the bottom of a ravine.
	10:10	Crowdsourcing site disabled, replaced with memorial message.
	10:10	TOTAL: 288 registered taggers identified 4,946 image locations.
	Download the full set of crowd tags as a Google Earth KML here.
Sun, July 29
100h	10:00	Bodies carried off mountain.
Mon, July 30
124h	10:00	Reports emerge about where the climbers were found. Location is 1,000 feet below the ridge where last tracks were identified from crowdsourcing.
		Ground photo shows face ascent, ridge descent and fall location.

LESSONS LEARNED

In a very short time, a dedicated community sprang up and searched a huge area of satellite imagery to find meaningful clues that successfully pinpointed the final tracks of the missing climbers. Once we acquired current imagery, the analysis time was incredibly fast with answers emerging in just a few hours.

Many different parties, spread across 3 continents, worked together to make this happen. In addition to the information created from crowdsourced image analysis, everyone involved felt a sense of empowerment, bonding over having an action which made an impact. Future deployments will include a community forum where taggers can post comments, discussions, photos and anything relevant to the search.

Our sincere thanks and appreciation go to:

• Everyone at DigitalGlobe and GeoEye for their immense effort and incredible response in making archival imagery available and tasking new collects of the mountain.
• The ground team in Peru for their dedication and determination in leading the search.
• The crowd of 288 who came together to help us search through the imagery and also donated to help the families fund the rescue mission.
• Ben and Gil’s families for their feedback and for allowing the community, many of whom knew and loved the climbers, to help with the search.

CONCLUSIONS

Gil and Ben are two of the finest climbers we have known who embodied the spirit of the mountains with every thing they did. Tomnod cares deeply about applying our technology to help in future search and rescue scenarios. The lessons learned from this tragedy will teach us to be faster, smarter and successful next time.

Gil and Ben pose on the Palisades Traverse (photo: Brad Wilson)

Our friends at DigitalGlobe and GeoEye have been incredibly helpful in providing imagery and support. Now people can quickly get to work looking through the imagery. Within a day we were able to get WorldView-2 to snap a picture and put it on Tomnod’s GlobalInsight platform. Now climbers, readers, technologists, people who care — and you — can help by donating time (or money to support the air search).

To help, go here: http://tomnod.com/gi/peru/

People are already spotting tracks and even the three members of the ground rescue team.

What looks like the three rescuers identified using crowdsourcing in the satellite image

So far, hundreds of people have contributed 2,421 tags (and counting). I am convinced that if there are clues in the satellite imagery, we will find them regardless of how subtle they are. We are currently collecting the crowd’s consensus points and following up on the leads to generate a collection of locations to direct the ground and air support teams in Peru.

Each of the points represents a clue in the imagery that people have spotted.

Amazingly, lower on the mountain we see the trail begin to appear as people contribute tags along it.

Take Action Now!
http://tomnod.com/gi/peru/

It’s a simple idea: Twitter + Geo.  What are people saying and where are they saying it?  These are basic questions, but getting the answers is surprisingly difficult.  In this blog post I’ll talk about some of the shortcomings with the Geo layer in Twitter’s API offerings, and what we built to get the functionality that we needed.

At first glance using Twitter’s built-in “geo” functionality seems pretty straightforward, but dive into it and you’ll soon realize that there’s a lot to be desired.  In fact, Talking with my good buddy Charles at GNIP, catching a stream of tweets coming from a geographic area is notoriously difficult, and there’s no clear good way to do it yet.

Twitter’s Built-in Geo-tags

The first thing you might try when trying to put tweets on the map is looking at the “geo” field in tweets returned from the Twitter Search API.  Go ahead, try it out:  http://search.twitter.com/search.json?q=blue%20angels .  Look at the “geo” fields in the returned results, and you’ll see that exactly zero tweets have embedded geotags.  In our own internal testing, we typically see that only about a quarter to a half a percent of tweets actually have embedded geotags.  I’ve never seen it above 1%.  That’s a lot of data flying around without a home.

Search API Location Information

Twitter’s search API allows you specify a location (in lat/lng) and a radius that those tweets should originate within, and it goes by both actual embedded geotags in tweets as well as the “location” field that people fill out in their Twitter profiles in free-form fashion.  Problem solved, right?

Well, not quite.  There are still a couple of problems with this:

If you’re monitoring many keywords, the number of searches you need to perform (number of keywords you care about times the number of locations you care about) starts to really add up and you quickly run into API limits trying to track it all.

The second problem is that Twitter’s Search API location search has been, and remains buggy and inconsistent.  Here are a few links to some of the issues that I’ve been following on this front:

https://dev.twitter.com/discussions/3360 – General Location Bugginess

https://dev.twitter.com/discussions/4004 – Geocode search results fall out of specified radius

https://dev.twitter.com/issues/98 – Issue #98: Geocode search results fall out of specified radius (allegedly fixed as of 2012-05-17)

https://dev.twitter.com/discussions/4003 – Geocode search volume lower than expected

https://dev.twitter.com/issues/141 – Issue #141: Geocode search volume lower than expected (not resolved)

But check this out: Perform this geo-search out in the Nevada desert:

http://search.twitter.com/search.json?q=&geocode=38.8,-118,10mi&rpp=100&result_type=recent  and all results fall completely outside the radius (as of June 19, 2012).

In fact all the results are from completely bogus locations (check the examples returned above).  Clearly it’s not fully fixed yet.

Streaming API Location Information

Well maybe the Twitter Streaming API offers a solution? The main problem with the Streaming API is that you can filter by keywords OR filter by location, but not both at the same time.  Here’s the official word from the Twitter documentation:

https://dev.twitter.com/docs/streaming-apis/parameters#locations:

Bounding boxes do not act as filters for other filter parameters. For exampletrack=twitter&locations=-122.75,36.8,-121.75,37.8 would match any tweets containing the term Twitter (even non-geo tweets) OR coming from the San Francisco area.

So we run into a similar usability problem as with the Search API: we need to either search by location then filter on our own by keyword, or search by keyword and filter ourselves by location.

Twitter: Geo Enhancements “Not on the Radar.”

So there are some problems with the way things are now, but what is Twitter doing about it?  At the recent Big Boulder Conference, Twitter said that working with geo was “not on the radar”.  They’ve got other things on their minds, probably like monetization, for one thing.

Do a search on Stackoverflow and nobody has yet come up with a clear solution:  http://stackoverflow.com/search?q=%5Btwitter-api%5D+location

The Path Forward: Tomnod’s Geo API

So what’s the solution?

We’ve been grappling with this problem over the past 18 months, refining our strategy for putting tweets on the map for a while now.  Among other things, we’ve built a Halloween map (http://www.utsandiego.com/halloween-map/) and a Holiday map (http://www.utsandiego.com/holiday-map/) for the San Diego Union Tribune, an Irish Times St. Paddy’s Day map (http://www.irishtimes.com/topics/st-patricks-2012/paddytimes.html), and a host of other tweet-mapping deployments.

Like any good hackers, we figured there MUST be a better way.

Also, like any good hackers, we decided to build it ourselves

So we built the tweet geolocation API that we wished existed 18 months ago when we began our journey into tweet-mapping.

Let me introduce it to you: It’s the Tomnod GeoAPI (http://geoapi.tomnod.com).  I’ll have more info on how it works and what exactly it does in a future blog post, but for now please feel free to kick the tires and try it out.

Cheers!

-Nate

Since I was among the Twitterati, I took it upon myself to record the best observations and insights via my own Twitter stream. This resulted in by far the most tweets I’ve ever sent in a 2-day period — but there was so much to talk about!
Here are my highlights of Big Boulder:

The first panel featured 2 leaders from Twitter: Ryan Sarver (@rsarver), Director of Platform and Doug Williams (@dougw) Partnerships. They shared some staggering stats:

I also learned about Twitter’s new Embedded Tweets product that I’m using to embed all the tweets in this post!
In the most bitter irony, Twitter went down for over an hour — right while the Twitter guys were on stage!

A recurring theme at Big Boulder was Big Data: how to collect it, analyze it and take action from it.

Next up was @JoeFernandez, CEO of Klout who gave a few juicy, behind-the-scenes insights:

With all this tweeting, I’m expecting my Klout score to skyrocket!

The last panel of Day 1 was also the most interesting. @Moeed Ahmad, Head of New Media at Al Jazeera, @RumiChunara from HealthMap at Harvard Medical School and my friend Katie Baucom (@katieb3505) from NGA spoke about how media, health and government are using social data for social good.

Day 2 kicked off with the exceptional Brad Feld giving an inspiring vision of how to create a startup community.

Given the job he’s done and the success he’s had in Boulder (e.g., with the Foundry Group, TechStars and investments like Gnip), we should all take note! I invited Brad to share his vision with San Diego entrepreneurs next time he’s on the West Coast.

Some of the toughest questions of the conference were directed at the only people in the room wearing suits: the guys from Facebook!

It was great to catch up with @MylesSutherland from ESRI who’s working with Tomnod to bring our crowdsourcing to their tools and data. Myles and his colleague SJ Camarata talked about Tomnod’s favorite topic: geo data.

Social Media Analytics

The conference wrapped up with a discussion of Social Data in the Markets with the always-engaging @HowardLindzon of StockTwits as well as Richard @Tibbetts, CTO of StreamBase who shared some great insight on working with the government.

As well as all the great speakers, I was really glad to meet some old friends and make some new colleagues, including:

• John Lucier (@trigorilla) from DigitalGlobe.
  John and DG are great partners of Tomnod. We spent a lot of time discussing how Tomnod will put human eyeballs on all of DG’s pixels: 240 billion of them every day!
• Megan Costello (@megcos) from Crimson Hexagon.
  Megan is working with me and the UN Global Pulse initiative to use Crimson Hexagon’s powerful tweet-mining to monitor social trends in Indonesia.
• Katie Baucomm (@katieb3505) from the National Geospatial Intelligence Agency.
  Katie never fails to impress and show how the stereotype of government being old, slow and non-innovative is completely wrong.
• @SeanGorman from GeoIQ.
  Sean and his team at GeoIQ are an inspiration to Tomnod: they show how to combine great technology with massive geo data to develop successful products for government and commercial customers.
• @SophiaBLiu from the U.S. Geological Survey.
  As custodians of the National Map and creators of Did You Feel It?, one of the best examples of geospatial crowdsourcing, Sophia and her colleagues at USGS push the boundaries of what can be done with geo data.
• @HowardLindzon, his wife and Chris Corriveau from StockTwits.
  Great to meet another awesome company from America’s Finest City!

I’d like to end by sending some huge thank-yous to:

• Everyone from Gnip who put on a fabulous conference.
• The people of Boulder who hike, talk, think, cook, drink, play, party — and startup — like few other places on earth.
• Joanna & @CharlesInce, true friends of Tomnod, for their endless hospitality, comfy couch and for letting me give Zoe her bottle!

Tomnod and our friends at DigitalGlobe were invited to present our recent work crowdsourcing imagery of Damascus at the Big Data Ignite session. John Lucier began with some staggering statistics about DigitalGlobe’s vast amount of imagery: 2,500,000km² is photographed every day – an area the size of Greenland! Then I delivered Tomnod’s vision to put human eyeballs on every pixel.

Below is a video of my talk together with John Lucier from DigitalGlobe.

All the talks in the session dealt with the “rising tide” of massive data. From moving, storing and processing exabytes of data to intelligent analytics and human-computer interfaces that extract intelligence from the information, the challenges and opportunities were thought-provoking. The Ignite format — fast paced 5 minute talks where each slide advances automatically — made for entertaining talks, although it was a challenge to fit everything in!

Bottom line: as geospatial data keeps growing, the industry is ripe for innovation … and people are getting after it!

Passive crowdsourcing relies on being able to gather information produced as a result of the existing behavior of a crowd of users or customers. If we include sales data analysis in this category — everything from supermarket loyalty cards to Google analytics on clickthrough rates — passive crowdsourcing is even older than active crowdsourcing. One of the most famous online examples of passive crowdsourcing is the pervasive reCAPTCHA which supplies websites with images of words from books and requires visitors to decipher the contents to prove that they are in fact human. By entering the correct information, users are unknowingly helping to digitize books. The project has so far transcribed the entire New York Times archive and has been acquired by Google to further digitize traditional print content.
 

Information + Location

To realize the power of passive crowdsourcing, many companies have sprung up and are using social media to track trends in various verticals. Tomnod adds a new dimension by using our Geo-Location API to add location information to social media content. For example, to measure the efficacy of an advertising campaign one can look at the change in sentiment and number of mentions on Twitter. Similarly, for the Republican primaries we deployed our algorithms on Twitter data and were able to predict results around 7 daysbeforethepolls. In this case, the application is not asking users to report information; rather passive crowdsourcing listens to what is being reported in a public forum. The wisdom of the crowd is already out there, we are just tapping into it.

If you build it, will they come?

Active crowdsourcing applications tend to emphasize ease of use to attract a crowd. By reducing the barrier to entry, the application is accessible by a larger crowd and the data collected is more accurate. As we continue to design these user-facing systems, we should take inspiration from passive crowdsourcing where the barriers to entry are so minimal that users do not even feel like they are working.

Three lessons from passive crowdsourcing:

1) When crowdsourcing applications are fun it makes it easier to recruit a crowd and their results are often more accurate. What if instead of farming corn on FarmVille users were actually completing a microtask contributing towards a project?
 

2) The crowd is talking — we need to listen. The crowd has wisdom beyond the direct clicks we ask them to do. Gather as many metrics as you can from the crowd by tapping into social media or their actual behavior in your application.
 

3) Go to where the crowd is. Instead of trying to attract your own crowd to your application, it is wise to tap into existing social networks and places where people already congregate.

 
 
Woot woot.

Greetings and Happy Friday!

The answer for last week’s puzzle is the Olympic Stadium in London. This stadium seats 80,000 people and be the primary site for the 2012 Olympic games. The most recent publicly available satellite photo of the stadium was taken in 2010, hence the unfinished appearance. For fun, try logging into Google Earth and finding the Olympic Stadium in London. The timeline toolbar will let you see the land on which it was built from 1945 to 2010. You can watch the Stadium rise up right before your eyes!

Now, here is this week’s Where in the World challenge:

Hint 1: This image was taken in a country that was once part of the Roman Empire.

Hint 2: Humans have occupied this site as early as the third millenium BCE.

Happy Hunting!