Google

Time for some random thoughts about Google and what they might be up to long term that could justify a 200 dollar stock price.

I for one am holding out hope that Google will turn out to be the M$ slayer. Well that's not really true. I would be surprised if M$ ever really dies.... But then no one thought IBM would be so far out of the PC OS business in the early 80's either. How could Google do this ? Well they are definitely taking strides to interconnecting people in ways that are probably long over due. They are making formally hard things easy... like blogging for example. And no writing a blog was never very hard for someone technically adept, but Blogger makes it as easy as a web e-mail system and yet allows for a great deal of customization if you so desire. It is a great example of Google's philosophy. Make things possible and stay out of the users way, and give it away to bring them in the door so that you can put context sensitive ads around them.

Many people have theorized that a Google OS may be in the offing and I for one happen to agree. I think before long you are going to see a Google Portal application that at least at first will be built around a browser concept on steroids. It will be a glue architecture for connecting all of Google's offerings into a single coherent package. Ultimately it will further blur the line between local and network storage and begin unifying the internet as a persistent environment. I think that ultimately they are going to have to get into the provider game and they are going to do the same thing they always have. Give it away and provide such a good service it attracts enough people to make money on unobtrusive ads.

How could they do this? Well They are buying a great deal of dark fiber which has tremendous data capacity. A true Network OS with standard programs running as web applications is going to take a pretty good pipe to the users... what better way to provide it than being able to determine the pipe people use? My off hand guess would be fiber running to last mile wireless nodes... and I mean wireless on a massive scale. Fiber lines can run at Terabytes per second rates and that can provide fat pipes to a lot of people. If you could make a line of sight (similar to radio or cell phones) fiber linked wireless system you could probably blanket most metro areas pretty damn easily. Yes it would cost a lot of money but that is something Google can leverage right now. On the plus side Fiber is highly under used and thus very cheap right now. Moving massive amounts of data around is their specialty and if they provided a usable universal wireless broad band access for free they could fricken guarantee people will buy the needed quantities of client hardware to tie into the system.

To put it another way. Google has the money to buy the fiber. They have the money to buy/develop the wireless equipment. If they are going to do this it would be fairly easy to talk a hardware manufacturer to build the needed client cards based on the fact they were providing it. If it works then in the end they make money for the hardware providers to provide people with the ability to connect to their system. They make money by having people on the system the same way they make money by people going through their search sight. In a sense it is the same thing they do now just on a larger scale. And it would drive more data through their systems for indexing which just makes their core function that much more capable.

If on top of this it will almost have to require you to access it via portal software... or better yet a portal OS that is a purpose built front end that leverage the power of the local system and provide access to the powerful network back end. This way you could wind up with some seriously cheap systems in a hurry... E-machines that work. Your machine would be a login on a portal system away. Similar to the way Google Mail is available wherever you go and have a net connection. Imagine if your desktop was the same way... including access to materials stored on your home machine like it was local ? Your backup capacity would be the industrial level provided by Google in addition to any private means you wished to take on your own. And that would in general be far more than most people do now. This wouldn't take away using the system like you do now necesarrily... but if done right it would just make it obsolete. It also would provide the means to impement needed security patches.... ie it would be similar to going from a 'confederate' system of networking to a 'federal' one.

Sound far fetched? Well Google is already working on the personal storage level that would be needed. 2.5 gb per gmail account and climbing. More and more uses for that storage space are coming online courtesy of blogger and picture sharing with Hello. The conversion to digital formats of library contents. How far fetched is a web service Open Office suite ? One with a constantly updating set of grammar rules culled from all users, one which automatically links to quoted texts or contextually similar information ?. Thin client front ends to a set back end. IE windows faces version and hardware compatibility nightmares. What if they had the barest requirements for access on the front end and had control of the programs provided on the back end ? The result should be far more stable IF you could rely on the connection. IE my system doesn't do the number crunching... all it provides is input and a place to output. In other words putting to practical use the concept of 'The network is the computer'. On top of this they have a SETI style CPU cycle system which could be put to use sharing capacity among the systems hooked to the network.

To make an idea like this work they are going to need pipes similar to sustained HD data transfer rates. That means a 10mbs up/down pipe at the least preferably with some burst capacity up around 100mbs. Then all you have to do is keep network latency under 10ms and viola... you don't even notice most of the data you are accessing isn't local. The performance would be very comparable to a local system. Wireless tech is already at 54mbs. Cell phone systems in Japan are already there. The fiber capacity is already available here. All that remains is an OS, and putting the system together. I give it 5 years max at which time either we will already not be able to recognize computing when compared to today... or well on the way. Google is the trailblazer cause they are the only ones with the balls to provide the access for free and know that if they get the herd to go through them they can make money in the process. So long as they can make it make money it will happen. I don't see this happening as a pay for access. At least not on that fast a time scale. If you make it negligible cost (50-100 dollar access card) to access it the adoption will be all but instantaneous by the people that matter. Once the pendulum swings that will be all she wrote for anyone that isn't riding the wave. At least that is what I think now. Perhaps next week will bring some new news.