Google has just announced that it will completely reshape the presentation of its search results. Instead of listing websites where people can find the answer to their queries, it will present data, links, pictures, etc. from its own databases, some 500 million "items". The redesign is said to represent the biggest change in search for the last five years, not just conceptually but also in terms of business impact.
The goal, of course, is to make sure that web surfers remain on Google properties instead of clicking away to another site for answers. Internet companies all over the world already complain and regulators can add the move to their list of Google features to investigate from an antitrust perspective. The announcement is also a subtle tactical move, just one day before Facebook's IPO. It reinforces Google's positioning as the owner of the "knowledge network" as opposed to the "social network".
It is hard to anticipate what the new 'product' will really be but there are a few important pointers. First, it will make connections between the different information elements. At the moment, search provides a list of sources but does not really categorize them. Knowledge Graph will and may even link the sources to one another. People (not least Tim Berners-Lee) have always argued for the, so called 'semantic web' to replace the traditional web. Google's Knowledge Graph is a fine approximation of the idea. Another aspect of the service will, hopefully, have to do with formatting. We may get a more standardized presentation of the information, which is quite important for rapid processing. Will Google look more like Wikipedia then? I think that the answer is 'yes'. The difference is that the voluntary human element in building knowledge will be taken out of the equation and be replaced by a few algorithms.
And now for some fun: I always thought that Google is building a giant brain, the cells of which being living web pages constantly built and animated by human beings. Is it possible that this brain will one day wake up to consciousness? It is hard to believe that this will not happen, actually. Maybe the alarm clock has just gone off....
Search This Blog
Friday, 18 May 2012
Wednesday, 25 April 2012
SEC considers civil case against Egan-Jones, an independent CRA
Egan-Jones is a relatively young, independent credit rating agency (CRA), one of few such institutions where investors pay for the agency's services not the institutions whose securities are being rated. As I have written many times before, this difference is important as it has been shown that the conflict of interest so fundamental to traditional large CRAs (like Moody's, Standard & Poor's, etc.) has been one of the reasons for the recent financial crisis and previous stock market bubbles.
I am not sure what the charges are (according to the April 20, 2012 issue of the WSJ they concern the provision of misleading information about the number of ratings provided and the amount of time Egan-Jones performed such ratings) but the possible consequences may essentially mean the (temporary) suspension of Egan-Jones' license as a CRA.
As we have no details about the evidence it is not possible to decide whether the SEC is right or wrong. Nevertheless, I find the case extraordinary. Despite broad criticism of the leading CRAs and their obvious role in the financial crisis no case was ever brought against them. Conflict of interest together with the lack of competition have been clearly identified as the main issue with the CRA industry, yet the SEC never really went after any of the top three firms, which together control the ratings industry in a tight oligopoly. In this context, making such a strong move against a small, barely 5-year old, independent agency that prouds itself not to be plagued by the fundamental problem of conflict of interest is a major statement and casts doubt on the SEC's willingness to reform the industry. We'll see -the SEC may well have a case. But given its behavior over the last decades in this matter, and in particular, its incapacity to foster competition I see little hope for fundamental change.
I am not sure what the charges are (according to the April 20, 2012 issue of the WSJ they concern the provision of misleading information about the number of ratings provided and the amount of time Egan-Jones performed such ratings) but the possible consequences may essentially mean the (temporary) suspension of Egan-Jones' license as a CRA.
As we have no details about the evidence it is not possible to decide whether the SEC is right or wrong. Nevertheless, I find the case extraordinary. Despite broad criticism of the leading CRAs and their obvious role in the financial crisis no case was ever brought against them. Conflict of interest together with the lack of competition have been clearly identified as the main issue with the CRA industry, yet the SEC never really went after any of the top three firms, which together control the ratings industry in a tight oligopoly. In this context, making such a strong move against a small, barely 5-year old, independent agency that prouds itself not to be plagued by the fundamental problem of conflict of interest is a major statement and casts doubt on the SEC's willingness to reform the industry. We'll see -the SEC may well have a case. But given its behavior over the last decades in this matter, and in particular, its incapacity to foster competition I see little hope for fundamental change.
Tuesday, 3 April 2012
On the media by the media
Kaifu Zhang, a doctoral student at INSEAD has highlighted a great article in the LA times that laments about the opposing perspectives that various media outlets gave on Trayvon Martin's tragic death: see
The article is a great testimony of classic media bias with quite surprising levels of divergence in the expressed views. But it is not just that. It is also a great testimony of agenda setting: media's capacity to focus hundreds of millions on a particular issue.
Monday, 2 April 2012
Social publishers: a new paradigm for online media
Many of my recent readings of the business press give the impression that a consensus seems to emerge among professional content providers (also called "content mills") - online newspapers (e.g. Huffington Post, The Business Insider), online tabloids (e.g. BuzzFeed) or blogs. The consensus is: "the thing to maximize is Facebook Likes!". This is interesting because it acknowledges a shift in the way people consume or "use" online media content. Under the previous paradigm, the idea was to build a portal where people spend time because the portal represents a large bundle of their online information needs. The bundle is designed to maximize the time spent or page views on the site, which then provides higher advertising revenues. Customer acquisition often meant search-engine optimization with the objective to become the "go to place" for a certain topic or fashionable buzz word.
The new paradigm sees the consumer as a social player on the Web rather than someone searching for news, entertainment or other information. People use social networks to promote themselves with an inherent need to build "relevance" in their community. Much more attention is given to the generation of content than to the consumption of content. In this context, the goal is to provide raw material for people to share. The content generated by online publishers - increasingly called social publishers - is an invaluable source for people to find material to share. If we manage to grab the attention of our peers and generate buzz around a piece of funny or touching content or news, we promote our relevance and eventually our social status. But, it is hard to do this with the pictures of our dog or cat or the random experiences that we report on our life. Breaking the news on a good story or a gossip just buys us so much more relevance.
Clearly, if this is the purpose of content search on the Web then the social publisher's objective is to generate as many "Likes" or positive feedback as possible with the posted content pieces. This will make the content pieces viral in social media (mostly Facebook) with the potential to reach a huge crowd. Each visitor spends little time on the social publisher's page but there will be many of them. Moreover, this viral phenomenon might also work for a new form of advertising where instead of showing banner ads to people, brand-specific content (that entertains) is shared between members of the community. There is evidence that ads seen in a social context (e.g. recommended by a friend) work much better than ads "on the periphery" of other content.
But, is it true that our social media use is essentially based on our 'narcissism' with the goal of promoting ourselves? Aren't we just trying to stay in touch and interact with our friends? Research indicates that our inherent tendency to compete for relevance is in many ways the essence of our humanity. The newest theory (by Jean-Louis Dessales) on the origins of human language - arguably one of the top candidates to define our uniqueness as humans - convincingly suggest that speech has evolved to improve individuals' capacity to build coalitions. Speech is a medium to advertise the person in an eternal competition for relevance with other members of the community. More relevance provides us with more capabilities to build useful coalitions. In this sense, similarly to our jokes at a party, our Facebook posts are 'performances' with the aim to build social capital (and at the risk of losing it). Social publishers pretend to help us by providing the raw material.....
The new paradigm sees the consumer as a social player on the Web rather than someone searching for news, entertainment or other information. People use social networks to promote themselves with an inherent need to build "relevance" in their community. Much more attention is given to the generation of content than to the consumption of content. In this context, the goal is to provide raw material for people to share. The content generated by online publishers - increasingly called social publishers - is an invaluable source for people to find material to share. If we manage to grab the attention of our peers and generate buzz around a piece of funny or touching content or news, we promote our relevance and eventually our social status. But, it is hard to do this with the pictures of our dog or cat or the random experiences that we report on our life. Breaking the news on a good story or a gossip just buys us so much more relevance.
Clearly, if this is the purpose of content search on the Web then the social publisher's objective is to generate as many "Likes" or positive feedback as possible with the posted content pieces. This will make the content pieces viral in social media (mostly Facebook) with the potential to reach a huge crowd. Each visitor spends little time on the social publisher's page but there will be many of them. Moreover, this viral phenomenon might also work for a new form of advertising where instead of showing banner ads to people, brand-specific content (that entertains) is shared between members of the community. There is evidence that ads seen in a social context (e.g. recommended by a friend) work much better than ads "on the periphery" of other content.
But, is it true that our social media use is essentially based on our 'narcissism' with the goal of promoting ourselves? Aren't we just trying to stay in touch and interact with our friends? Research indicates that our inherent tendency to compete for relevance is in many ways the essence of our humanity. The newest theory (by Jean-Louis Dessales) on the origins of human language - arguably one of the top candidates to define our uniqueness as humans - convincingly suggest that speech has evolved to improve individuals' capacity to build coalitions. Speech is a medium to advertise the person in an eternal competition for relevance with other members of the community. More relevance provides us with more capabilities to build useful coalitions. In this sense, similarly to our jokes at a party, our Facebook posts are 'performances' with the aim to build social capital (and at the risk of losing it). Social publishers pretend to help us by providing the raw material.....
Tuesday, 27 March 2012
Freedom and the Internet
The Jasmine Revolution that has affected many of the Arab countries showed how big a role the Internet in general and social networks in particular can play in promoting political freedom. Social networks promoted the free flow of information and also helped people organize spontaneous demonstrations against oppressive regimes.
Yet, on second thought, it is not entirely clear that the Internet automatically guarantees free speech. The fact that user-generated content is recorded on Internet servers makes it easier for authorities to track contributors and punish them if the promoted information and ideas are not supporting the regime. The Chinese government, for example, has just introduced regulation that requires people to setup their accounts on microblogging sites and social networks with their true identities. Foreign sites (e.g. Twitter) are banned and local sites (SinaWeibo, Tencent Weibo, Renren, etc.) are constrained by such regulation. Enforcing the law is not a trivial task (the trade of bogus IDs is very liquid) but still the recorded information makes it easier for authorities to track down individuals who disseminate undesirable information.
But the government can be more subtle in the control of information by actually influencing it. In a recent Economist article ("The power of microblogs", March 17, 2012, p.55) it is reported that "... [Chinese] government agencies, party organs and individual officials have set up more than 50,000 weibo accounts" to influence the public debate about current issues and controversies that are massively affected by the rumors circulating on the Internet. This environment is particularly interesting and it is not clear who - the uncoordinated individual citizens or their coordinated government counterparts - are at a disadvantage in building consensus from the noisy cacophony humming on the Web. Maybe, with a sophisticated government the Internet is a more dangerous threat for freedom than traditional media....
Yet, on second thought, it is not entirely clear that the Internet automatically guarantees free speech. The fact that user-generated content is recorded on Internet servers makes it easier for authorities to track contributors and punish them if the promoted information and ideas are not supporting the regime. The Chinese government, for example, has just introduced regulation that requires people to setup their accounts on microblogging sites and social networks with their true identities. Foreign sites (e.g. Twitter) are banned and local sites (SinaWeibo, Tencent Weibo, Renren, etc.) are constrained by such regulation. Enforcing the law is not a trivial task (the trade of bogus IDs is very liquid) but still the recorded information makes it easier for authorities to track down individuals who disseminate undesirable information.
But the government can be more subtle in the control of information by actually influencing it. In a recent Economist article ("The power of microblogs", March 17, 2012, p.55) it is reported that "... [Chinese] government agencies, party organs and individual officials have set up more than 50,000 weibo accounts" to influence the public debate about current issues and controversies that are massively affected by the rumors circulating on the Internet. This environment is particularly interesting and it is not clear who - the uncoordinated individual citizens or their coordinated government counterparts - are at a disadvantage in building consensus from the noisy cacophony humming on the Web. Maybe, with a sophisticated government the Internet is a more dangerous threat for freedom than traditional media....
Saturday, 24 March 2012
Credit Rating Agencies
Last week has been rich in news about Credit Rating Agencies (CRAs). Both the Financial Times (March 23, 2012) and The Economist (March 17, 2012) had articles on CRAs commenting on how the industry is being reshaped after the scandals triggered by the global credit and fiscal crises. I have followed this industry for over a decade now but I keep being surprised how little progress has been achieved in fixing it.
The Financial Times reports Europe's efforts to regulate the big three CRAs (Standard & Poor's, Moody's and Fitch). European governments are angry because CRAs have downgraded their debt. It doesn't really help the regulatory effort, of course, that its leader is Michel Barnier, a french carrier politician who knows little about economics and probably even less about information markets. His idea of regulation is to put a straight jacket on CRAs, e.g. by making them liable for their decisions or forcing issuers to rotate CRAs. Obviously, simply introducing more competition (e.g. reducing the barriers to entry) would be far more effective in making the industry more careful with the ratings, without mentioning the cost effectiveness of implementing such 'deregulation'.
The Economist reports on the evolution of CRAs in India. It praises the local industry wondering what might explain its success given that local CRAs use the exact same model observed in the US, i.e. exhibiting massive conflict of interest, which, of course, has been in big part responsible for the global financial crisis. One argument advanced is that in India, agencies have been better regulated by the local financial supervisor (SEBI) and India'd central bank. Of course, we all trust Indian government agencies in doing a good job at overseeing and regulating the economy. Is this a joke - really?
The other argument raised is that the CRAs in India have done a better job at diversifying their businesses. I am not sure I understood this argument. Essentially, incompetence and conflict of interest are going to be spread to other sectors of the economy and we should cheer about this? The article mentions one agency moving into hospital ratings, for example...
When evaluating threats to the industry, The Economist mentions that new entrants might represent a danger. Presumably, these would provide soft ratings to gain market share. Of course, we don't expect that conflict of interest will lead to biased ratings from the incumbents who are comfortable in controlling the market (?). Again, I am lost in the reasoning.... Reading between the lines, the picture gets even grimmer. India has 6 licensed agencies with three large 'local' agencies (CRISIL, CARE and ICRA) leading the pack. Reading on, we discover that CRISIL, the largest (worth $1.3 billion) is 52% owned by Standard & Poor's. In turn, Moody's owns 29% of ICRA. Given the uncertainty surrounding India's economic institutions and the devastating climate of corruption the apparent health of the local CRA oligopoly is rather depressing.
The Financial Times reports Europe's efforts to regulate the big three CRAs (Standard & Poor's, Moody's and Fitch). European governments are angry because CRAs have downgraded their debt. It doesn't really help the regulatory effort, of course, that its leader is Michel Barnier, a french carrier politician who knows little about economics and probably even less about information markets. His idea of regulation is to put a straight jacket on CRAs, e.g. by making them liable for their decisions or forcing issuers to rotate CRAs. Obviously, simply introducing more competition (e.g. reducing the barriers to entry) would be far more effective in making the industry more careful with the ratings, without mentioning the cost effectiveness of implementing such 'deregulation'.
The Economist reports on the evolution of CRAs in India. It praises the local industry wondering what might explain its success given that local CRAs use the exact same model observed in the US, i.e. exhibiting massive conflict of interest, which, of course, has been in big part responsible for the global financial crisis. One argument advanced is that in India, agencies have been better regulated by the local financial supervisor (SEBI) and India'd central bank. Of course, we all trust Indian government agencies in doing a good job at overseeing and regulating the economy. Is this a joke - really?
The other argument raised is that the CRAs in India have done a better job at diversifying their businesses. I am not sure I understood this argument. Essentially, incompetence and conflict of interest are going to be spread to other sectors of the economy and we should cheer about this? The article mentions one agency moving into hospital ratings, for example...
When evaluating threats to the industry, The Economist mentions that new entrants might represent a danger. Presumably, these would provide soft ratings to gain market share. Of course, we don't expect that conflict of interest will lead to biased ratings from the incumbents who are comfortable in controlling the market (?). Again, I am lost in the reasoning.... Reading between the lines, the picture gets even grimmer. India has 6 licensed agencies with three large 'local' agencies (CRISIL, CARE and ICRA) leading the pack. Reading on, we discover that CRISIL, the largest (worth $1.3 billion) is 52% owned by Standard & Poor's. In turn, Moody's owns 29% of ICRA. Given the uncertainty surrounding India's economic institutions and the devastating climate of corruption the apparent health of the local CRA oligopoly is rather depressing.
Wednesday, 29 February 2012
The currency of social media is INFLUENCE
Here is a great summary of the kind f practical research we have done in the domain of social network analysis or 'social network marketing.
It is by Zsolt Katona from Berkeley, one of the founders of Lynx Analytics
Subscribe to:
Posts (Atom)