Krishwords

In this guardian article, Nick Carr says

That would be nice, but I have my doubts. When we all have equally precise, equally up-to-the-second information on traffic conditions, the odds are that we’ll all respond in similar ways. As we all act in unison to avoid one bottleneck, we’ll just create a new bottleneck. We may come to look back fondly on the days when information was less uniformly distributed.

This is plain rubbish. The technology that informed people to move from the first bottleneck due to real time traffic information will also inform people about how traffic is diverging away from the first bottleneck. It will help drivers make an informed decision about their route and, in fact, an intelligent system may even help the driver route better based on instantaneous analysis of traffic divergence from the first bottleneck. The above statement is just nonsense. Technology doesn’t work selectively. Period. He shouldn’t be making such statements just to fill up a column that is due.

Duncan Riley of Techcrunch says

Google Reader’s share tools on the other hand republish full blogs post for all to read without obtaining permission from blog publishers. So-called link blogs in Reader already break copyright and in a small way undermine blogs and content creators. If Google offers a comment service on “shared” items they are in effect creating copyright infringing blogs; after all they’ll have chronological entries and comments so they’ll look like blogs, even if they don’t provide a fully customizable CMS.

Doesn’t this sound like Sam Zell’s complaint against Google News? I am not sure if Mike Arrington subscribes to this view of Duncan Riley. If he does, I would say that techcrunch is still in the Web 1.0 era while covering a lot about Web 2.0 companies. Isn’t Mike Arrington the guy who was pitching Digg as a replacement for New York Times? I wonder what he feels when his employee takes an entirely opposite stance when it comes to new social media technologies. Duncan dude, get over it. The way we consume media has changed a lot. The ideas about copyright has changed. As long as the original source is linked in the linkblogs, it is not a copyright violation. If linkblog is bad, techmeme is bad, digg is bad and we can extend such a logic to claim that the whole idea of social media is bad. It is not Google that should drop the idea. It is Duncan Riley who should drop the old fashioned thinking about media consumption. Times are changing dude. Being the torch bearer of Web 2.0 companies, it doesn’t look good to talk in Web 1.0 slang. Robert responds to Duncan’s comments here.

Electronic Frontier Finland has released a report that points to an interesting piece of data. The countries where corruption is high has voted in favor of Microsoft. Add to this the news reports about Microsoft employee who was caught offering compensation for partners who are willing to say “Yes” to OOXML. Sounds fishy to me.

International Organization for Standardization (ISO) rejected the fast-track approval of the controversial Microsoft-supported OOXML document format as an ISO standard in a vote on 2 September 2007. During the voting process the reputation of ISO as a dependable technical standardization organization was questioned. For example, in Sweden a Microsoft representative was caught offering to recompense partners for voting yes to OOXML. Also a sudden interest from countries like Ivory Coast to the OOXML issue has been found suspicious. We studied the relation between the corruption level and voting behaviours of the countries. We found that more corrupted the country is, the more likely it was to vote for the unreserved acceptance of the OOXML standard proposal.

Image Credit: effi.org

Robert Scoble kicked off a discussion today with his prediction that Mahalo, Techmeme and Facebook will beat Google. In the videos, he seems to be pretty convinced that Mahalo can beat Google. I disagree with Robert’s take on this. I had a few back to back tweets with Robert on the topic. In one of the tweets he asked whether several Mahalos can get together to beat Google. In this post, I am going to disagree with the thesis that “Mahalo can beat Google”. I am using Mahalo here but it applies to any human powered search engine with Mahalo’s approach to search.

Before I offer my views on the topic, I want to make the following clear.

• I completely agree with Robert that SEMs (Search Engine Marketers) are a pain and we need better search results.
• If, by any chance, Robert had meant that Google should be beaten in the search game, I am in full agreement with it too. We need to stop Google’s possible search monopoly and there should be heavy competition to force Google to continue innovating.

Having said that, I want to categorically state that Mahalo, in its current form, can’t beat Google. The reasons are many. Some of them include

• Human powered search cannot scale like machine based search
• Human powered search has built in bias in the system
• Human powered search will definitely lack in freshness
• Human powered search doesn’t mean much in non text search world like Image Search, Video Search, Book Search, etc.
• Human powered search is a ridiculous thing when it comes to desktop search.

There are many other reasons where Mahalo and other similar search engines will fail. However, these are the few reasons that just popped out immediately after watching Scoble’s video. Let us dig a little deeper on these reasons.

Human powered search engine in the Mahalo mould cannot scale like an algorithm based search like Google. Even if we assume that Mahalo grows like how Robert expects in the next four years, with 100K “volunteers”, it still cannot cover all topics in the world, it still cannot cover all languages in the world, it still cannot cover all the specifics in a particular topic, it cannot cover all the variations of a particular query, etc.. It might cover a search query “HDTV” much better than Google. But I searched Mahalo with queries “Dent on HDTV” and “Scratch on HDTV”, an hour back. It returned back saying there are no matching results. Even though it offers better results for HDTV, it fails miserably when you get into specifics on HDTV. Imagine the variations in queries on HDTV from all over the world, “HDTV with a spot”, “HDTV with a broken edge”, “HDTV and Bestbuy problem”, “HDTV with loose switch”, “HDTV’s remote not working”, “I just peed on HDTV”, etc. There is no way Mahalo can scale to take care of such specifics on even a single topic, along with everything else. Comparatively, HDTV is an easy problem to solve. There are much more difficult problems in this world which Mahalo cannot solve or may not bother to solve. Let us say I make a web page about a topic HDTV that contains a sentence like “not gone triple ribbon belt mother cook with mould acumen”. If someone searches for HDTV, my page may show up after the first 50 pages of Google results and not at all on Mahalo. But if someone searches for the sentence I mentioned above with double quotes, it will show up in the Google results and there is almost zero chance that it will show up in Mahalo (unless if my content is a killer content except for that particular sentence I quoted above). Now imagine endless new queries. Ask.com CEO claimed that 60% of their searches are unique. Even Google seems to have claimed that 20-25% of their total queries were not made before. Mahalo cannot even attempt to solve the search engine needs of all the internet users in this world, even if it scales like what Scoble hopes. It is like trying to build a 200 storey building without involving any machines and just with one brick at a time. It is just not humanly possible. Scaling problems will ensure that any attempts to create a completely human powered search engine is a failure.

To highlight my arguments, I am adding the picture of the result I got in Mahalo when I queried “mod_security rules for SQL injection attack”. You can see how badly it failed for this query. The scaling of Mahalo in the next 4 years, which Robert describes, may answer this particular query well. But, there are billions of other queries which may throw up similar results even after four years.

Human powered search like Mahalo has inbuilt bias which is difficult to manage. Jason feels that this bias can be managed with stuff like transparency, ethics codes, hiring people well, firing bad people quickly, early warning systems, etc. He claims that they take a Neutral Point of View (NPOV) on stuff like abortion, 9/11 and George Bush. I agree with him that with stuff like transparency, ethics code, etc., you can take a NPOV on few things like the queries he has quoted. Bring in diversity of queries and the different points of view attached with it. Now bring scale these number of queries. There is no way you can maintain NPOV consistently on all the topics and various queries associated with these topics. Jason even suggested the idea of Mahalo Ombudsman and included Jeff Jarvis as one possible candidate, which he rejected immediately. Even if he makes it ombudsmen instead of ombudsman, the scaling problem will ensure that the bias will stay in the system. Add to this the issues like embargo. For example, the embargo on Iran might prevent Jason to hire guides from Iran and this may result in inferior Mahalo results as far as topics related to Iran are concerned. But, an algorithmic engine like Google can just crawl the websites in Iran and have better results on the topics related to Iran.

Human powered search engine like Mahalo cannot keep the content fresh. Google crawls the web more frequently to keep its contents fresh. Now, I think, we can even inform Google about how often it should update our websites and this helps Google to update its results with fresh contents. There is no way Mahalo can keep the content fresh on all topics and related to all queries. The scalability problem described above, will make sure that the Mahalo content is stale except for a smaller percentage of terms.

I think it is pretty obvious that human powered engines like Mahalo cannot do much in the areas like Image search, Video search, Book search, etc. Add to this, the search in other languages, reverse phone number look up, etc. I don’t even have to talk about how ridiculous it will be to consider human powered search when it comes to desktop search.

Robert ignores the longtail concept when he talks about human powered search engines replacing the algorithm based ones. The very concept of longtail came into existence due to the phenomenon called internet. Under such a scenario, betting the future of internet on something that uses the hits and ignores the longtail (remember his discussion on social fabric? He talks about bumping the results up based on the number of people who trust a particular person (Michael Arringtons and Robert Scobles of the world)). I do agree that Google uses pagerank which can, in some ways, be likened to the hits in the longtail concept. However, a well structured query in Google will also get results from the longtail of web pages whereas it will not be fetched in the social fabric approach of Mahalo. This is a very important point and I just hope I have put it forward clearly.

So what is the solution to the search engine problem? Is Mahalo doomed for failure? What about 100s of Mahalos? Well, I am not an expert in search engines to offer a solution to the problems, in the current day search engines. But I strongly believe that human based search engines like Mahalo cannot replace (or beat) algorithm based search engines like Google. My solution is still algorithmic based. My hunch feeling is that the future search engine will be a collection of vertical search engines which may use Robert Scoble’s trust concept to optimize the search results. As I told Robert in our twitter conversation, Mahalo can, at the very best, be a Google optimizer but not a Google beater. Vertical search engines, with Mahalos in the front for optimization, may offer the much needed solution to our search engine problems. Humans cannot replace machines in the search engine world. Mahalo may serve well for a niche market but it is not a Google beater. At best, it can be termed as a Google optimizer.

What is your take on Scoble’s thesis that Mahalo and other trust based human search engines will beat Google? Do you agree with him? If not, why do you think human search engines cannot beat Google? What do you think will be the Google killer?

PS: It is 2:15 AM. I will correct any mistakes in the morning.

I have been following the controversy that arose based on what Dave Winer told about Jason Calacanis‘ presentation at Gnomedex. Apart from the merits in Dave Winer’s argument about lack of value for users from Mahalo, I found one thing a bit funny. I am following both Dave Winer and Jason Calacanis in the blogosphere. I notice their styles and their philosophy. I find it a bit strange that Jason, who doesn’t mind speaking his mind even if it is hurtful to the opposite party, is upset when someone else, who might have a higher position in the hierarchy of the blogosphere, speaks their mind against him. I find this a bit hypocritical, with nothing personal against Jason. It might help Mahalo, if he finds a way to incorporate Dave Winer’s valid suggestion. Why should people come to Mahalo from Google if it can’t offer a value to users like Google does. Getting search results is just a matter of getting your queries right and getting smarter to avoid spam. You cannot base your company just on the search engine spam. You need to offer more value than what Google offers to its user base. In fact, I am thinking of writing a post on what I see as Mahalo’s disadvantages. But it is for another day and it may even be irrelevant for Jason. But Dave Winer’s suggestion about making Mahalo a platform and allow users to benefit from it, is very important if Mahalo has to gain some traction.

If you are curious to know how the offices of various tech startups look like? Check out Office Snapshots. Looking at the offices of various startups, you can get a feel for their character. Thanks to the person who is doing this, you get to have a better idea of these companies you read in Techcrunch and other places.

Link thanks to Marshall

Google LatLong Blog says

Google Maps now lets you know how long a drive might take in rush-hour traffic, for a limited set of metropolitan areas. Of course, if there is an accident, the drive could take even longer. But more likely than not,

Tags: Google, Google Maps, Traffic

Los Angeles Times has this story that talks about how second life marketers are closing shops.

Four years after Second Life debuted, some marketers are second-guessing the money and time they’ve put into it.

“There’s not a compelling reason to stay,” said Brian McGuinness, vice president of Aloft, a brand of Starwood Hotels & Resorts Worldwide Inc. that is closing its Second Life shop and donating its virtual land to the nonprofit social-networking group TakingITGlobal.

CNET Australia has this article that talks about the Web 2.0 culture and how it resembles the dotcom culture

It’s time to take off that uncomfortable suit and put on those comfy jeans you left at home during the dot-com crash. And while you’re at it, grow that tidy corporate haircut out, let your facial hair run wild, and visit your local tattoo parlor so you can show off some visible ink.

The word is out. IT rock geeks are back in demand and stereotypical “dot-com” culture (and smell) is back in vogue. Managers are again in a bidding war to compete with their rivals and new Web juggernauts like Google to retain their best employees by offering a laid-back environment to benefit staff moral, retention and productivity.

Enter Web 2.0 work culture, the future of yesterday.

We have (re)launched TechBizMedia. We are going to use this blog to offer our analysis on various topics like Web 2.0, Social Media, Bioinformatics, etc. We are planning to cut down the hype and offer a balanced analysis on what is happening in the tech world today. I strongly recommend you to subscribe to this new blog. I will be one of the authors there. Feel free to buzz me if you want to offer any suggestions for this new blog.