Navigating the Future: Take Me to Bob

Google has just announced a free turn-by-turn navigation app for Android 2.0 in the US (Radar post). Google Maps Navigation relies on Google’s own mapping for routing you. As with many navigation devices you can search Business Listings. However, they are also including data not traditionally available to navigators. In the promo video Google demonstrates that you can ask to be taken to “The King Tut exhibit”. GMN will determine that it’s in Golden Gate Park and route you. This is “because it is connected to the internet it is using all of the latest information on the internet.”

This is huge. To be able to request implicit destinations based off of realtime information is something that has never been available before. What new queries will be available to us because of this? Google has a lot of data. How much of it can be assigned a location? Lots. There are millions of KML files out on the internet. Here are some of the useful queries

“Take me to Bob Smith” – If Bob is your friend on Latitude then Google Maps Navigation can take you to him. If Bob moves then GMN could even re-route you. I wonder if they will enable the chase scenario.

“Drop me off in time for the #48 bus” – Google knows the public transit schedule. So not only can it drop you off at the nearest stop, it could drop you off at the stop that will ensure the shortest multi-modal trip.

“Show me homes under 500K in Capitol Hill” – Via Google Base, Google has real estate information (it has had neighborhood data for quite sometime).

“Take me to my next appointment” – If you use Google Calendar and you accurately fill out the location field then this is a snap.

“Take me to the nearest Winter Coat Sale” – Using Adsense for Google Maps, GMN can easily lead you to local sales.

“Take me to the bar my friends go to the most” – Using Social Graph API and the new, experimental Social Search to tap into Foursquare, GMN can determine where you friends go, aggregate their destinations and lead you to their favorite watering hole.

“Take me to the largest event” – Using a combination of Latitude and its new access to the Twitter Firehose (which will soon include locationRadar post), Google can determine where people are.

“Take me on a tour of the top 10 historical sites here” – Using Wikipedia Google can determine what the sites are and where you should be taken. Alternately, Google could take you on user-generated tour.

“Take me to the most picturesque place near here” – Several years ago Google bought Panoramio, a location-based photo site. Google can determine which place nearby has had the most photos of it taken.

“Take me on a tour of the site from Around the World in 80 Days – Google already geoparses many of the books it scans (just see this map). This routing is quite possible.

“Take me to the EPA’s protected sites” – Government data is becoming more available. This is just one possible governmental query. You could also ask to go on a tour of TARP fund recipients or Democratic donors.

Obviously not all of them will be enabled, but I bet that within a year some of them will be. What other scenarios can or should they implement?

tags: , , ,
  • http://tim.oreilly.com/ Tim O'Reilly

    I want to emphasize a point I made in my “Web Squared” paper (http://assets.en.oreilly.com/1/event/28/web2009_websquared-whitepaper.pdf), and have been hitting in a lot of my talks:

    the next stage of web 2.0, which I defined as the development of database-driven services that get better the more people use them, is the development of applications that depend on coordinated data from multiple databases.

    Voice activated search is one of these applications:

    1. Google has an online speech recognition database (which gets better the more people use it)

    2. Google has what John Battelle famously called “the database of intentions” – the likelihood of search intent, which you see every day in the form of Google Suggest. This helps them do better speech recognition, because they know what you are likely to be saying.

    3. Now they’re adding tighter and tighter integration with their location-related data assets (which will also get better the more people use them.)

    Microsoft is trying to acquire and control similar assets, but this doesn’t have to be just a two-horse race. There is another alternative, which could be likened to the design of Linux (or the internet itself) as a collection of independent cooperating systems. So, for example, NavTech (owned by Nokia) could be working with Nuance (which does speech recognition) to deliver similar apps. And of course a whole other dimension could be added by bringing in data from Facebook, etc.

    If I were any of these companies, I’d be thinking about how what can be called the “small pieces loosely joined” approach to the internet operating system we’re building can stand against the “one ring to rule them all” approach.

    With that context, go back up and read Brady’s post again, seeing just how many amazing data assets Google is assembling and bringing to bear in delivering its next generation internet applications. I love where Google is going, but I also think the future is better with lots of competition, so I’d like to see others figuring out how to go there as well. Few have all the capabilities that Google has assembled, but by working together via federation, there may be interesting alternatives.

    In any event, there is a major strategic watershed ahead for companies that don’t want to be taken out of play in the way that the navigation data providers and dedicated device providers are being taken out right now.

  • http://dudegalea.co.uk Pat Galea

    All good points.

    I would add a point that’s at a tangent.

    Google Maps is awesome when you have connectivity. Not so great when you’re offline, either out in the boondocks, in an urban Faraday cage, or cellular roaming without data.

    I’d like to see Google degrade gracefully in these situations, perhaps caching map tiles for the surrounding area whenever it can. So, for example, if you’re using hotel wifi, cache the map tiles for the surrounding 400 sq km at various resolutions so they’re available when you’re outside.

    I’d also like to see the Google search engine work offline to some extent. Perhaps it could cache local geographic results together with meta information allowing a Google app to return appropriate results without having to hit the net. (I envisage this an extension to Gears, but with more intelligence added.)

    So my phone would get the full Google functionality, search and maps, when it has data connectivity. And it would degrade nicely when the network drops out. Ideally, for a suitable set of queries, I wouldn’t even notice whether I was connected or not.

  • http://epeus.blogspot.com Kevin Marks

    “Take me to Bob” already works, as long as you’re both using Latitude – I used it to find people at OSCon…

    Better if they integrate it with existing calendar and friend location feeds from FireEagle, Foursquare, Brightkite et al than get everyone to use Latitude though.

  • http://twocroissants.wordpress.com Bertil Hatt

    Can I say how both excited and terrified I am about that information? Excited because it can be way more powerful then what you describe — and terrified because, even with just “take me to my next appointment” we become so dependant on that machine it really starts to scare me. Am I wrong to insist on people knowing how to tell North, draw the map of where you are going, and have basic knowledge of how streets are numbered? I can’t imagine managing a appointement list is an obsolete skill now —I’m relieved, given my schedule— but I see too many car-drivers nose-down in their cellphone right now not to feel shivers watching this video and reading your comment. I must be wrong; I’d love to be.

  • Sam Penrose

    @Tim: one obvious conclusion from Brady and your comments is that “the web” starts to become an anachronism. Berners-Lee and friends derived it from sets links between documents, which doesn’t apply here. Do we really want to call these new tools and capabilities “web (anything)”? I’m not sure we want to call them “Internet (anything)” either. Per @Pat’s smart comments, the network not the magic. Per your “coordinated data from multiple databases”, the heart of the phenomena is something like the automated emergence of knowledge from the accumulation of data. Knowledge 0.1? Taught machines 0.1? A.I. 2.0? I lack a gift for the apt phrase here … anyone?

  • http://twocroissants.wordpress.com Bertil Hatt

    Can I say how both excited and terrified I am about that information? Excited because it can be way more powerful then what you describe — and terrified because, even with just “take me to my next appointment” we become so dependant on that machine it really starts to scare me. Am I wrong to insist on people knowing how to tell North, draw the map of where you are going, and have basic knowledge of how streets are numbered? I can’t imagine managing a appointment list is an obsolete skill now —I’m relieved, given my schedule— but I see too many car-drivers nose-down in their cellphone right now not to feel shivers watching this video and reading your comment. I must be wrong; I’d love to be.

  • http://www.ianpeter.com Ian Peter

    These data matching capabilities also give rise to a whole lot of privacy issues that are bound to come up fairly quickly. Some more queries that might be addressed with these capabilities include:

    * where is my husband and where is that old girlfriend of his?

    * is client X at work today?

    * just who is at bar x tonight?

    * use of real estate lists as theft targets

    I wonder where privacy legislation which typically addresses data matching intersects with some of these potential capabilities.

  • http://www.cloudsourced.com Christopher Osborne

    @Pat one of the main reasons for ditching Tele Atlas as a map data provider is to avoid licensing restrictions. Previously, you were not allowed to cache much map data on the mobile device. Now they are using their own data they can do anything they like. I would expect they will be offering exactly what you want shortly!

  • http://http:twitter.com/ambercoldren Amber Coldren

    Will be interesting to see the privacy issues … AKA “big brother syndrome”. Ian makes a great point.

  • Bertrand de La Chapelle

    Excellent remarks from Ian. Behavioral tracking ? You also can imagine, if there is data retention by the different databases, then the law enforcement agencies will naturally be interested in the collated data : was this person at that place at that time ? Or can someone provide an alibi in court by asking the provider to release the data concerning him/her at a particular date ….

    Many new situations can arise that would be worth exploring in the future – for instance at the Internet Governance Forum.

    Should the developers of such applications set up advisory boards gathering all stakeholders to define balanced terms of service incorporating those privacy issues ?

    Should a set of public policy principles be established to deal with the balance between the potential benefits of new services and the potential risks to privacy.

    B.

  • http://www.chuvakin.org Anton Chuvakin

    “Take me to the most picturesque place near here”

    If this doesn’t sound like AI, I don’t know what is…

  • http://phil.ashlock.us Philip Ashlock

    This also sounds a lot like the AIDA project at MIT:

    http://gigaom.com/2009/10/29/mit-is-building-terminators-nascar-cousin/

    http://senseable.mit.edu/aida/

    Unfortunately, the implementation of AIDA is completely car-centric.

  • http://www.mortgatecalculator.net Ajeet

    I must confess that most of this blog post made me get the creepy feeling that humans are on the brink of a scary lack of ability to live out of the “matrix.”