Friday, October 23, 2009

Google vs Microsoft, who will win?

High-tech companies are very exciting. Just like watching a game of baseball, where not much could happen for 7 innings, and then all of a sudden, a bad pitch, a good hit, and a stroke of luck could change the game immediately. Even big companies that's been winning game after game, could suddenly be toppled because of something as small as an obnoxious fan trying to catch a ball. IBM, Microsoft, Apple, Sony, Intel, AMD, Sun, all big name companies at one time or another, all went through phases of drastic change even within the few decades of operation. No company is immune, and they all need to stay vigilant to changes in the fickle market. And now two of the biggest tech companies have up'd the ante, and you better be at the edge of your seats because a homerun could change the game at any time. How does Windows 7 change things? Here's how it could play out:

How Microsoft could kill Google: 
Bing Bling
Google does a lot of things, email, maps, web hosting, documents, but they really only make money from two things -- adwords and adsense, both of which are tied to their search engine. The billions of dollars they are paying their engineers to have fun and develop cool web apps come from the bidding of keywords that is displayed on the search results page. If Google lose the search engine market or even just a chunk of it, they would suddenly have too many engineers with not enough money to support them. If this were to happen, say goodbye to all the Google benefits, people would start losing their jobs, and the culture would have to be drastically changed to defend their core business. The company would not be able to sustain the heavy investments they have in R&D, and would have to resort to being like Microsoft, where they would only enter a market after someone else have proven its worth. Microsoft can then deal the final blow because, you know, Microsoft is better at being Microsoft than Google is.

So, how would they pull it off?
To put it simply, create a beachhead. Some may hear that Bing search offers pretty good search results, but still not use it. Why change to bing when Google works great?

What they need to do is to give a reason to start using Bing. It doesn't have to give better results for everything, or try to replace Google immediately, but it needs to have a "wow, this is awesome" feature to use Bing for. If Microsoft developed Google Scholar, Google Images, or Google News before Google did, they would have a very strong starting point. From there, after users get used to typing in www.bing.com to do those specific searches, users would seriously consider replacing Google since they both solve the need for search. What they need is a means of attracting a core audience, and then find ways of spreading that audience to use their service for all other purposes. Specific and useful searches, but simple and easy to switch to their main service.

Will it happen?
Microsoft picked real-time search for their first beachhead. It certainly has potential to be big. However, I don't know about you, but I'm still confused about the whole twitter and update things in facebook. I'm not sure what circumstances I would need to be in for me to think in my head, "hmmm, what should I use to help me solve this problem? Oh, I know, real-time search with Bing!" But I would keep an eye out for any developments along this front.

I personally wouldn't have signed both Twitter and Facebook at the same time, since they are pretty fierce competitors. Rather, I'd probably try to sign an exclusive deal with Twitter, just because Facebook doesn't seem to like Google a lot. Still, though, props to the big MS for seeing this market and getting the deals through.

So... maybe. It could happen. It depends on whether people are right about the importance of web 2.0 for search and how Google will respond to this threat.

side note: 
How about make a 2d search result page, that offer the top results from each type of search, and the user can move through the results depending on which direction seems to match what the user is looking for the most? You can even eliminate the pagination delay with Ajax and make it like cool like iPhone.

How Google could kill Microsoft: 
Cloud OS Online Office
Microsoft, even though they have significantly diversified, their core business and most of their innovation still comes from their operating system, Windows. Without it, their relationship with computer manufacturers couldn't be leveraged, and their dominance would fade when they don't have enough money to throw at new markets to compete with the leaders there.

So, how would they pull it off?
Well, Google is one of only a few companies in the world that can pull off the creation of a real cloud operating system, as I described here. But honestly, that's still a few years away from coming to market.

An alternative is to make operating systems a sort of commodity, making people care for operating systems as much as they care about the type of tissue they have in the bathroom. Once that happens, it would no longer make sense to pay $200 for an operating system when you can get one for free, or another one that gives you a personality.

What do you think of when you're thinking about changing operating systems? Usually, two things -- programs and files. Right now, there are a bunch of web tools that can do what have been traditionally done only on the computer. But out of all of those, the most important ones are the productive tools that you use to make the presentations to the boss. He doesn't care which program you use or which operating system you have, just that you get it done. If you can just upload your presentation to Google docs in an easy manner (like dropbox), and be able to make it show up when you're ready to present, it wouldn't matter that you made the slides on your Linux box, and you're presenting it on a computer in the conference room with a mac.

Will it happen?
No. As much as I like Google, and as much as I don't like Microsoft, I really don't they are positioned well enough to take on Microsoft. Honestly, it feels like they've given up on Google docs. They haven't updated it in ages. For the longest time, I tried to like it, but it's a lost cause. Now that more and more people are creating pptx files that Google docs can't naturally convert, they've lost the window of opportunity.

Google creates a lot of mediocre products, sprinkled with neat-o features, but that is really not going to make a dent in the Microsoft kingdom. Luckily for Google, Microsoft also has to worry about Apple. They just might distract Microsoft enough that Google comes out with something that makes even more money than search. yaa.. I won't count on it though.

Conclusion
To me, it seems like Microsoft is playing it very smart in an Armani-suit-businessy-way, whereas Google is just playing around with promising platforms and cool applications without a focused strategy. They would seriously be in a bad position if they suddenly find Microsoft with a significant portion of search and their newest gizmo's like Android, Wave, and App Engine turn out not to be the money-makers they expected. Microsoft have played defense and offense pretty well, and they are poised to keep their position as the top player. Let's hope the Google engineers have been using their 20 percent time well.

And no, Facebook won't be able to replace either of them.

Friday, October 9, 2009

resume business cards

one liner: a business card that also has your resume on it

Especially in this economic environment, there are a lot of people looking for jobs. Sometimes you're at a gathering, and you find someone that would be interested in who you are, possibly a person from HR, or a manager, or an investor. It would be great if you could give him a business card, especially if it has your resume or your sales pitch or your business plan on it.

some options for how to do it. 1. put a detachable flash drive or sd card on it, but it might be expensive. 2. turn your resume into a qr code but the amount of stuff you put in is limited 3. print really really small, and offer a magnifying glass =P 4. link to your website, but the user has to type it in.

Any more ideas?

Thursday, October 8, 2009

Detecting DNA with nanomachines

DNA is small. Despite being the code that spells out how to make each of the billions upon billions of proteins that make up who you are, each base, or letter of DNA, is 2 nano meters wide, and 0.3 nanometers long. Even stretched out, 800,000 letters of DNA can fit around a single strand of hair. One difference in a million separates you from smelly Bob, and only a few changes in key places would give you cancer. So how would you read something so small, yet so important? Well, how about with tiny diving board?

In 2000, a significant work [1] from J. Fritz and the team at the IBM Zurich Research Laboratory presented a method that have been commonly used to create CPUs to sense biomolecules, including DNA, RNA, and proteins. They used semiconductor fabrication technology to create silicon cantilevers, which are tiny diving boards measuring 1 x 500 x 100 microns to create a platform where the materials would be detected. These cantilevers are still much larger than biomolecules, but they are small enough that they would be affected by them.

So to detect a specific sequence of DNA, they put complementary sequences stacked vertically on the cantilever. When a sample is placed on the device, the DNA from the sample would pair up and bind with those on the platform. The surface that was already crowded would now be even more packed, and the stress that is only on only one side of the cantilever causes it to bend, which can then be detected with high accuracy by a laser.

To test out their device, they created two such cantilevers, one with a 12 letter DNA sequence and the other with a 16 letter sequence and then they put two samples with the complementary sequence one after the other on the device. Of course, we cannot perfectly predict the reactions between the things in the sample and the DNA molecules on the platform, but with two such platforms, any difference would be significant. If the things in the samples bound to one platform stronger than the other, we can assume it was caused by the very specific DNA pairing. Sure enough, the laser signal showed the cantilever wobble and then stabilize with one cantilever bent lower after the first sample was added, and then wobble and stabilize with the other cantilever bent lower after the second sample was added. Even DNA sequences that differed by only one base pair caused a noticeable difference. Other than just attaching DNA sequences to the platform, they also tried proteins -- to see whether it interacted with another protein, and antibodies -- which can be made to bind to just about anything.

Microarrays and high throughput sequencing also offer scientists information on the microscopic world, yet both require probes and a complex procedure to perform the experiment. These micro cantilevers offer an unique advantage in its simplicity. Since the publication of the paper, people have proposed microfluidic devices that pump samples into a chamber with many cantilevers for detection, incredibly complex, but completely automated and miniaturized to fit just about anywhere.

We now know a lot about DNA, proteins, and other tiny molecules that makes a big difference in a living organism, but so far we haven't been able to detect these molecules cheaply and reliably to help with medical diagnosis. The devices that we do have are large and clunky. Soon, however, we will have laboratories on a chip that will be able to put all this knowledge to good use for everyone. And it is all thanks to a very tiny diving board.

[1] Fritz, J. et al., “Translating Biomolecular Recognition into Nanomechanics,” Science 288, 316, 2000.

Monday, September 14, 2009

material design?

http://www.techeblog.com/index.php/tech-gadget/phone-book-friction-unlocking-inter-connected-phone-books

could this be used to design new types of fabrics?

Wednesday, September 9, 2009

Online salesmen

One liner: provide an easy, low pressure mechanism for online shoppers to get assistance online.

There are many things that websites do well. But one thing that is hard to replace is the human interaction that are vital to any physical shop. Going to a car lot, few people go there to just look at cars and their brochures by themselves, they usually do that with the assistance of a car salesman, who knows the product, and can talk you though what you are looking for to find you the best match for the car. They answer questions, offer suggestions, make sales pitches, and create a pleasant atmosphere for shopping. To see how it makes a difference, go to any restaurant in Japan, where at the moment you walk in the door, they greet you with a hearty welcome, "Irashaimasae!", and knowing that the store is ready to serve you makes you want to patronize at the store even more.

Many online stores have good search functions, helpful faqs, or useful descriptions that go unused and unread. When a user doesn't find something before their patience runs out (which is really fast), they just leave the site to look for it elsewhere. Of course, this is bad for the site.

This idea is to allow instant messenging integrated with websites and the use of online salesmen to promote shopping on your site. Users coming to the site would be drawn to see an initial greeting from the salesmen, after which, the messenging window would put in an inconspicuous place that would not distract the user from normal website navigation, but still be in view and available for questions.

On the backend, each salesperson would be able to customize their own greetings, and chat windows would appear only when users reply to the greeting. Once in a chat, the salesperson would be able to see information relating to the user, such as the webpages that the user has been to, and the username along with the shopping history if the person has logged in. More likely than not, the user will ask about things that are already on the site. Therefore, it would be helpful to have a prediction system that will offer answers to the questions, where the salesperson would be able to use for their responses, such as pressing ctrl-1 to insert the first canned response suggestion. It should be made in such a way that maximizes the ability of salespeople to serve as many people as possible without compromizing quality.

One great thing about selling stuff online is the metrics that measure the activity on your site, and this would be no exception. Information about how much this system has helped your store can be clearly logged. Neat charts and pretty graphs can be made for users that have used the feature, and the additional revenue that came from the salespeople. The sales ability of the salespeople would also be measured, and commissions and bonuses can be given fairly. Depending on the measurements, the site owner would be able to fine tune any of these things, such as only offer the feature for customers that have already purchased X amount, make the chat window visible for only certain parts of the site where the margin is much higher, or assign different salespeople to different parts of the site. This feature would also help the site become more user-friendly, where the salespeople would be able to offer specific suggestions on where users gets confused or what most people are interested in.

Another great thing about web technology is that it can be done anywhere. These online sales people would be able to work at home, and probably at all kinds of different hours. This would be a position opened to anyone that would want to apply. Given a little training, it wouldn't be too hard to work in such a position.

[edit] just found out that this is currently being done (well) by olark.com

Friday, August 28, 2009

request for articles

one liner: in news aggregation sites, provide a voted request for for a certain topic

This one is simple, sometimes people are interested in certain things that no one wrote about, sometimes blog writers don't know what to write about. Make a way for readers to request information about a topic.

Wednesday, August 26, 2009

Ideas and Startups

Something just clicked for me today while reading posts about bootstrapping startups (starting from here), that startups are businesses and businesses are supposed to make money. Ideas are like ammunition. You need to execute the ideas to get value, where each one can provide you with a slice of the market. However, having more ammunition doesn't mean you win the war. These ideas are supposed to be like resources that a company has to go from one profitable business to another. This sounds blindingly obvious, but there has always been this misconception in the back of my mind that startups are there to prove whether an idea is good or not. This post is just a self-reminder that the business comes first, ideas have no value on their own.
[of course... we'll see about all this when I actually start my first startup]