Google Social Circle

Google labs is testing a very interesting new feature within the Google search results which lists and ranks content from people that have connections to your own social networks, websites, blogs, etc.   It’s called Google Social Circle and I think this approach has a lot of potential.

Google labs writes:

We’ve taken steps to improve the relevance of our search results with personalization, but today’s launch takes that one step further. With Social Search, Google finds relevant public content from your friends and contacts and highlights it for you at the bottom of your search results. When I do a simple query for [new york], Google Social Search includes my friend’s blog on the results page …

Filtering the massive oceans of content is what Google has been doing so effectively for some time, but the social media explosion has created a new kind of relevance Google’s basic ranking system has not been taking into account.    The content of trusted friends and associates is often going to be more relevant to us than that of, say, internet marketeers in a foreign country.     If, for example, my pal has travelled to Morrocco I’m going to trust his stuff – and probably be more interested in it – than information from strangers.    Google Social Circle will incorporate that relevance into the search results, and I think by doing this they may succeed where Facebook and Twitter have pretty dramatically failed.    Facebook’s search system and layout – in my experience – makes it very hard to search for information.  It can even be difficult to find a person you know, let alone find content they have created that is relevant to your search.    Twitter lists are something of a step in the right direction of targeting for relevant information, but Twitter search is severely lacking and I don’t even know if they they are particularly interested in providing the kind of contextual content mapping Google is testing with Social Circle.

Another interesting – some would say sinister – aspect of this approach by Google is to create internet environments filled with “trusted online information sources” that have been endorsed by different networks of friends.    Clever use of the data flowing in will allow Google to better screen sites based on human input, which is much harder to spoof than manipulations commonly done as part of aggressive “Search Engine Optimization” tactics.

The Social Circle reminds me of an advanced version of “del.icio.us”, a tagging and bookmarking service aquired (and largely abandoned?) by Yahoo a few years ago.  Delicious allowed users to tag and label sites and content, creating link lists of things relevant to them and giving them the ability to share these links with others.    By automating that process and using their brilliant search algorithm to slice and dice individual information, Google has pushed us one step closer to the holy grail of search – a system that shows us exactly what we want/need to see even if we cannot clearly state exactly what we want or need.

Google Researchers Make Image Recognition Breakthrough

Google research has announced they may have reached a significant milestone in image recognition.   In a demonstration and paper they’ll present today at an imaging conference Google will:

…. begin with an unnamed, untagged picture of a landmark, enter its web address into the recognition engine, and poof — the computer identifies and names it: “Recognized Landmark: Acropolis, Athens, Greece.” Thanks computer.

Although they explain this is not a new Google project, the implications of a very robust computerized imaging are very significant.   I’m not clear how this research intersects or relates to the facial recognition work of Riya and other companies, but as *hundreds of billions* of images pour onto the web from all over the world and as artificial intelligence systems such as the autonomous self driving vehicles of the Darpa challenge evolve, image recognition is certainly a very key element of the innovations that are driving computing forward.

Human information processing is primarily driven by visual interpretations and cues, so this may be considered something of an Artificial Intelligence milestone.

Wolfram Alpha Search. It’s no Google.

Reporting:  Joe Hunkins

Early hype suggesting that new search engine Wolfram Alpha could be a possible “Google Killer” quickly shifted to a focus on Wolfram’s new approach to search, which they call “computational”.     Although Technology-Report had early access to the program it is now open to all here:  http://www.wolframalpha.com

Although I’ve only spent a short time looking for inspiration at Wolfram Alpha, I’d have to say I could not find any answers where I felt Wolfram would beat out a Google search combined with some quick scans of the listed resources.     Wolfram’s promise was to deliver “the answer” to complicated questions but it seems to work well only for the kinds of information it appears they have already sliced and diced into packages, and I’m not clear it even beats out a Wikipedia entry when searching for data like states or countries where a packaged approach to the information is best.

A quick comparison of Wolfram’s answer to “New York” vs Google’s vs Wikipedia’s

In a case like this I’d argue Wikipedia is the clear winner, giving the user extensive information and links to more.   Google second with good lists, and Wolfram a distant third with very limited information given the wealth of data online.

For students creating notes (or papers to hand in!) Wolfram may provide some great tools with its unique organization schema, but for most internet researchers and browsers I think Google has nothing to worry about here at all.

Google Chrome: It’s a very good browser, so why don’t we use it?

When Google Chrome launched several months ago I think a lot of folks assumed they’d be switching to that browser, which uses several excellent innovations to enhance online navigation.     Google even issued a nifty comic book to help explain the innovation, and blogs were buzzing for weeks with mostly neutral or favorable reviews.

So what happened?    Why is Google Chrome market share so small compared to Firefox and IE?

The first reason of course is simply  … habit ….   It’s very hard to get people – even innovative online folks – to change from one good application to another.   Contrary to a lot of silly suggestions the Internet Explorer browser was not broken and even though FireFox has slowly been gaining market share it is clear that the rapid demise of IE was greatly exaggerated.     I use FireFox but I’d hardly say it’s dramatically superior or even all that different from IE.

Although it’s hurting Google Chrome, our habituation works very well for Google in the search sphere where people tend to use Google for search without even testing against other engines – that game is over and until we see a major new semantic search innovation Google’s likely to be the search of choice for years to come.

Interestingly Google Chrome really does “feel” different to me and on balance I liked the differences, yet like millions of other onliners who loaded up Chrome I did not switch over and rarely use that browser now.  I know one of my concerns was the uncertainty that still surrounds Google’s treatment of the data I indirectly share with them using Google products.   As a regular user of Google search, Gmail, blogger, and more watching Google both become dominant and also struggle to maintain their legendarily high online revenue I do worry that Google has too great a potential to become “the boss of me”.

Still not sure what’s up with Chrome, but as with many things internet it’s good to head over to Matt Cutts’ blog to get a very well informed opinion.     Matt is one of a handful of Google’s veteran search engineers and writes what for many is the key blog discussing issues relating to search, especially Google search.

Matt’s Five Reasons to Use Google Chrome

Matt’s Ten Reasons Not to Like Google Chrome

Hmmm – I don’t think it’s fair to use the 5 vs 10 math here, but maybe Matt’s on to something.   As creatures of habit we tend to settle in to the familiar and with the new we quickly look for things that bother us.   Google Chrome may in fact be the best browser, and I think I’d want to take the Crhome side in a debate even though I’m not using it, but not sure if I’ll be able to break the old browser habits.    Will you?

Artificial Intuition a key to AI?

Convergence08 was a great conference with many interesting people and ideas. Thankfully the number of crackpots was very low, and even the “new age” mysticism stuff was at a minimum. Instead I found hundreds of authors, doctors, biologists, programmers, engineers, physicists, and more clear thinking folks all interested in how the new technologies will shape our world in ways more profound than we have ever experienced before.

My favorite insights came from Monica Anderson’s presentation on her approach to AI programming, which she calls “Artificial Intuition“. Unlike all other approaches to AI I’m familiar with Anderson uses biological evolution as her main analogs for conceptualizing human intelligence. I see this approach as almost a *given* if you have a good understanding of humans and thought, but it’s actually not a popular conceptual framework for AI, where most approaches rely on complex algorithmic logic – logic that Anderson argues clearly did not spawn human intelligence via evolution. Yet Anderson is by no means a programming neophyte – she’s a software engineer who has researched AI for some time, then spent two years programming at Google and then quit to start her own company, convinced that her AI approaches are on the right track.

Anderson’s work is especially impressive to me because as someone with a lot of work in biology under my belt (academically as well as corporeally) it has always surprised me how poorly many computer programmers understand even rudimentary biological concepts such as the underlying simplicity of the human neocortex and the basic principles of evolution which I’d argue emphatically have defined *every single aspect* of our human intelligence over a slow and clumsy, hit and miss process operating over millions of years. I think programmers tend to focus on mathematics and rule systems which are great modelling systems but probably a very poor analog for intelligence. This focus has in many ways poisoned the well of understanding about what humans and other animals do when they … think… which I continue to maintain is “not all that special”.

Anderson’s conceptual framework eliminates what I see as a key impediment to creating strong AI with conventional software engineering – ie having to build a massively complex programmable emulation of human thought. Instead, her approach ties together many simple routines that emulate the simple ways animals have developed to effectively interact with a changing environment.

Combining Anderson’s approach to the programming with the physical models of the neocortical column such as IBM Blue Brain would be my best bet for success in the AI field.

Live from Convergence08 Conference

Mountain View, California: Convergence08 conference.  Hundreds of people are gathering here at the Computer History Museum in Mountain View “Bringing Life to Big Ideas”.    The focus is on four core technologies and how they will change the world dramatically in the coming decades:   Infotech, Cogtech, Nanotech, and Biotech.

I’m especially interested in hearing from Peter Norvig, Google researcher and the guy who – literally – wrote the textbook on Artificial Intelligence.

So far the organization of this conference is very impressive though it’s not clear how many are attending.    The group keynote begins in about 20 minutes.

Online Advertising Primer from Google

Alex at Google is explaining Google’s decision to buy DoubleClick and in doing so offers one of the best advertising primers I’ve run across.   People seem to have an enormous difficulty understanding why contextually targeted search ads tend to be a lot more effective than offline advertising and this will help them.  Online banner ads are probably just as crappy as offline ads.   This is a key reason Google has done so well – they dominate the contextual ad market while Yahoo has struggled to deliver a similar quality product.  Yahoo now has a good contextual ad product, but some think it’s too late for Yahoo to capture a big part of this market.   My view is that the game has only begun.

Jimmy Wales on Charlie Rose

Jimmy Wales, founder of Wikipedia, discusses his Wikia search projectand the internet. He’s the chairman of Wikia, Inc. He thinks it’ll be 2-3 years before they have a robust product.

“Democratic, participatory” search project.
“Google, Yahoo, Ask” have similar, proprietary and closed search. He wants to break up the idea that a few companies should be so dominant.

Making search ubiquitous. He thinks Google may not have problems with WIKIA because they can keep matching up ads, advertisers, and buyers as they have been.

Wales thinks Facebook made the right decision to turn down Yahoo’s billion+ offer for Facebook, calling it an “interesting gamble”. “He’s a pretty sharp guy” (Zuckerman), and Wales thinks that unlike Myspace, Facebook is doing right by the customers. Notes increase of spam and advertising intensity of Myspace.

Wikia major initiatives: Search, Reference Works for humor, opinion, sports. 66 languages plus a “Klingon language” project. “Roll this revolution” into many other areas. What makes the internet great is that it’s a “global platform for people to share knowledge”. Keeping it “open” appears to be a key guiding principle for Wales, and his admirable efforts at Wikipedia support his sincerity in that mission.

Wales suggests that Firefox is the best browser, primarily due to features that he sees as the result of the open source development model that created Firefox.    He says that monopolistic activity by Microsoft has slowed innovation, but feels that Google is a friend of Open Source.     Wales recounted telling Bill Gates at Davos that Microsoft search is so bad people are switching away from it as the Vista default, and suggests that he’ll have fun trying to build a better search than Google with Wikia.

Google and Privacy

Here is a nice post from Google about their new policy to anonymize search info from users. Like many I have been critical in the past of Google and others for storing this information with little regard to who owns it or saying what they’ll be doing with it.     Yahoo and MSN do not (yet) have similar policies so I think Google can rightly claim a higher road since they have also been the one who has fought Government attempts to nab search data.   (I have mixed feelings about that since, unlike folks like Battelle, I fear commercial abuses  more than I fear the Government will use my data in illegal and harmful ways.

Twitter and SEO

Interesting.   My Chico the Wonder Dog SEO experiment is yielding some unexpected results.    A tweet about this is now higher in the ranks than the original blog post page.

Chico the Wonder Dog has been trading places with another Chico the Wonder Dog.   That post is much older and may have more incoming links since that guy seems to spend more time posting about his dog than I do, though based on my quick analysis of this and a few other cases I think it indicates that Google looks carefully at the rate of link growth, and if it slows they tend to put back the “old, tried and true” page in favor of the newcomer. This makes sense from an anti-spam perspective although in Chicos particular case it probably does not yield the top dog.

However, the Twitter reference rising to high seems really surprising because Twitter posts are generally small and insignificant (as it is here).  I’m surprised Google ranks these at all, let alone makes them competitive with meaty postings.  Perhaps Google has elevated “social media” in some algorithmic fashion though my guess is this is a defect that will be corrected – ie Twitter is structured in a way that links to these posts from many Twitter people and this is messing up the Algo’s handing of this insignificant material.    If I’m searching for “Tesla Coil”, let along pretty much anything of any relevance, I hardly want a bunch of Twitter posts!