Search- is interface the new battleground?
Posted by Jeffrey on Feb 19, 2008
I was thinking recently about all the work being done around “natural language” search with several startups (notably powerset.com and textdigger.com) looking to make everyone’s search experience less frustrating. Let’s face it; despite all the work going into algorythms behind the scene, I don’t think that search today is significantly different than 5 years ago.You type in your word combos into the box and hope and pray that something relevant comes back in the first 10 entries on the page. Or you can repeat with a slightly different boolean combo. And still hope and pray…
But maybe its not simply in the initial search terms that determine how successful the search for information will be. Maybe if we acknowledge that its really really hard to get back the results we want on the first try, that the way to develop a better search experience is how you engage with the results to refine that search.
That’s why I was interested in Quintura and more recently in silobreaker.com. In each case they look at related items in a search visually in order to expose connections to the key search term. Don’t think an item is related? Just get rid of it either by clicking the ‘x’ or in the case of silobreaker, dragging it to the trash.
It makes refining the search more intuitive but also exposes linkages between items by proximity and size. Which helps you understand what the internet google Quintura/Silobreaker thinks of you.
Note: Another visual search engine was launched in Beta called SearchMe. Techcrunch covers it here
Interview with Amazon’s Jeff Bezon
Posted by Jeffrey on Apr 17, 2007
Jeff Bezos was one of the keynote speakers at the Web2.0 conference and in his eagerly anticipated talk he took the opportunity to highlight some of their initiatives to grow to the next level by providing outside businesses access to their world-class systems and technology. This approach would have seemed completely backwards about 5 years ago. Why provide potential competitors access to key technological assets and speed their time to market? Jeff tried to outline some of their thinking by kicking off his talk using their Amazon S3 internet-enabled file storage system; the same system they use internally for their e-commerce site Amazon.com.Amazon S3 has grown from about 800m objects stored in the system in July 2006 to over 5 Billion objects today. To illustrate some of the benefits of using the service, Bezos put the site Blueorigin (his effort to build a space vehicle) and all the video for the site on S3 so that they didn’t have to worry about scaling. This ended up being a prescient move as the traffic spiked when someone posted the site on Digg.
In fact on that day in January after getting Dugg, the website and media objects on S3 responded to 3 million requests. There were countless downloads of the video leading to transfer of 758Gb of data in one day. They did this without a contract or any sales contact; it was all done self serve by filling out some forms on the site. And the best part of it was that for all of January it cost $304 dollars, with most of the charges coming for that one day!
On S3s busiest day they were serving 1 billion requests per day and almost 16k requests/ second. Interestingly enough users include a company we may be familiar with; Microsoft. Secondlife and Powerset are also among the users.
Another example of what Amazon is working on with their services involves a service called EC2 (EC2 stands for Elastic Cloud.) If for instance your business needed to process many media files from avi to mp4, you could use their EC2 service to process all these files, which traditionally has been a very processor intensive task.
What would happen is that the raw media files stored in S3 pass messages to a queue service in EC2. EC2 finds a pointer in S3 to that avi file and starts the processing into an mp4.
To maintain quality control when the queue grows long, the EC2 can spawn new EC2s to clone itself to increase throughput of the service. So in effect you have an on-demand virtual server environment, which you only have to pay for when used. After processing the objects the virtual servers are redeployed to other tasks so your business does not have to pay for all those processors 24/7. You can use what they call “pay-per-drink” pricing in order to only use what you need at that time.
But a key question is how did Amazon get from a bookstore to an infrastructure provider? They took learnings from how to scale and deliver Amazon.com to other applications. Once they had this knowledge, they realized that they could then leverage this and open up this on a pay-per use basis so that companies didn’t have to reinvent the wheel every time they wanted to create an application or business.
The question then comes to mind whether Amazon is ‘traditional’ Web2.0 business? They have grown up alongside other trailblazing Web 1.0 companies but in contrast to the newer Web2.0 companies, they also have many hard offline assets including huge warehouses. But even thought they have 10 million ft2 of fulfillment space they are trying to again leverage their processes and infrastructure to outside players to create an ecosystem.
They reason that physical movement of goods won’t go away in next few years. So they will take the high cost services that businesses up to now have been forced to build or partner to get and allow businesses to exploit Amazon’s fulfillment network. You could think of this as the ‘programmable warehouse.’
You would simply hook into Amazon’s system and send a message that they should expect to receive items you will send over for storage in their warehouse. And the cost? Only 45 cents per cubic foot of storage per month. Then you can send a message to pick those things and Amazon would send them to an address you specify.
So they look at the future of Web2.0 a bit differently; they don’t just look at what will be the big disruptors, but they also put dollars into what they think won’t change. Interestingly, they feel they can build a strategy around what won’t change rather than what might.