Actually, reader, I’m a little tired of all these search posts. But new things keep happening and this one is compelling enough to note. I really miss Google’s notebook feature (actually a lot of people do). It was like BasKet for the Web. It sounds like Yahoo! is about to launch a new app called Search Pad that will be like Google’s notebook but with a teensy bit of intelligence.

This sounds like a right combination. If the search engine can be intelligent enough to figure out that you’re doing some sort of research and then help you with an easy-to-use note-taking, organizing system, fantastic. But if it becomes even more intelligent and can offer even more useful things than just archiving notes, that would be a powerful assistant.

There is some nice potential here. I wonder if Yahoo! will take advantage or underwhelm. Either way Google please take note, your competitors’ efforts to improve how people use the search results they get are becoming more sophisticated and intelligent. Will Wave make up for the loss of notebook?

Its preview side categories, recent searches, etc. have been done before by other search engines including Yahoo, Ask, Google, etc. A lot of Bing’s approach seems to be various ways of presenting search refinements, again not new (alltheweb.com the past days of teoma, and others). Its powerset stuff inside I suppose is important–helps Bing against some of the others (hakia.com) out there, but in the end I don’t seem to get better results than Google gives me and I still find Google’s interface preferable.

In practice Bing doesn’t appear to live up to its release claims of helping me make decisions. I don’t see much to justify Microsoft calling Bing a decision engine, so that must have been more marketing hubris than anything.

If the game is going change perhaps that will come from more interesting experiments like Wolfram|Alpha or an alternate company. Then again, Google continues to innovate and change the search game better than its challengers do.

“…search engines do a decent job of helping people navigate the Web and find information, but they don’t do a very good job of enabling people to use the information they find… Bing is an important first step forward in our long-term effort to deliver innovations in search that enable people to find information quickly and use the information they’ve found to accomplish tasks and make smart decisions.”

This goes hand-in-hand with the idea I was calling for when I said we need to ask How do I assemble knowledge from the information I find? In Erick Schonfeld’s detailed article on TechCrunch, he shows a number of examples of how this better guided search, with a decision-oriented result can work. It seems Bing will provide a lot of smart contextual information around searches to better identify what people are looking for and deliver results that a person will find useful for taking some sort of action.

I said before that we lack intelligent authors to assemble a structured knowledge for acquisition from the content of the Internet. The search engine should morph or augment itself into something that will provide that or come close to providing that (or else be usurped by an alternate application).

Microsoft is making a lot of a statements about changing the search game. Considering the launch of Bing occurs in such close proximity to the launch of Wolfram|Alpha, I think it’s fair to say that there is a real chance people are recognizing the need to change the search game in this direction. Wolfram|Alpha calls itself a computational knowledge engine and makes a point of differentiating itself from search engines. Wolfram|Alpha explains that

“You enter your question or calculation, and Wolfram|Alpha uses its built-in algorithms and growing collection of data to compute the answer. Based on a new kind of knowledge-based computing…”

From what I can see in Microsoft’s Bing and Wolfram|Alpha, neither have fully accomplished the game-change that is needed. However they’re each on the path. It’s clear that Google has a wealth of apps and interconnected information, which they’re subtly making visible and more-and-more intelligent over time. If you spend any time flipping back and forth between Google’s variety of apps, you start to see some interesting connections that overlay search results. Rather than launch an entirely new system, I think Google is evolving itself in this direction. I like that Bing and Wolfram|Alpha have become quite public, I think it will push Google more and it’ll be fun to see how these companies finally deliver on CASAA.

Search engines initially answered our question of “How or where can I find the information I want?” but in indexing the content of the Internet and providing access, they created a much more troubling problem. That question tends to overshadow another question, which is equally if not more important, “How do I assemble knowledge from the information I find?” That question will be solved by computer-assisted shallow atom assembly, which I think may be a new significant stage of Internet-related development.

Let me open up the problem a bit more.

The ubquity of “search” (whether through an Internet search engine, individual site search, etc.) has given us a more burdensome onus to seek out and gather the right shallow atoms for the knowledge we need. Search engines both enable and testify to this. Although they haven’t come close to solving my second question they’ve become slightly more sensitive to it: search results are not simply lists pointing to likely sources of information sought. Search engines offer more context for the results. They encourage things like voting on relevance, they offer information related to the results but tangential, for example searching for a company’s name will often also offer news about the company, stock symbols, addresses, display ads about related products and services, etc.

Search engines increased the pressure on individuals to find and assemble shallow atoms of information into the knowledge message or narrative answering their need for knowledge. Search engines provided us with functional access to the immense universe of discourse that is the Internet but it’s all based on the unadulterated results of computation. We thus lack intelligent authors to assemble a structured knowledge for acquisition from the content of the Internet. The search engine should morph or augment itself into something that will provide that or come close to providing that (or else be usurped by an alternate application). That will be a great advance.

The Continuum

I’ve been saying that we have a more shallow atoms now and have adapted our attention spans suitably for the labour we must undertake to acquire knowledge. Here’s what I mean.

Although we’ve always had, more or less, shallow atoms for knowledge acquisition, I think that the availability of these has increased in our Internet-oriented era. I see the deep unit and shallow atoms as points on a continuum of methods for knowledge acquisition. Event without electronic means, we have shallow atoms. For example, a poster plastered to a construction site might announce a rally for a political cause. It would likely contain just a few words or phrases about the cause with some pertinent information on the time and place of the rally. Someone seeing this atom would still need to acquire information from other sources in order to develop sufficient knowledge for understanding the issue. Magazine articles too, these lie in a more middle ground of the continuum. They’re relatively short like an atom but they’re structured more closely to a unit.

Virtually, we’ve created so many new types of shallow atoms, not to mention facsimiles of the physical ones I just mentioned. Take the Web concept of linking, which just encourages the shallow atom mode of defining information. Because we can include a link within one shallow atom to another, there is no need to deepen the individual atom. An atom that references something requiring more information includes a link to one or more atoms to satisfy that requirement (and those atoms may include links as well).

Considering this continuum is essential to considering the changes in our methods of knowledge acquisition. Whereas I’m asserting we’ve had various forms of knowledge from the continuum for a long time, I also think that the technology of the Internet and that which branches from it have immensely increased the quantity, forms, and occurances of shallow atoms. The impact on people acquiring knowledge is that while perhaps we maintained a certain type of balance in terms of analysing, interpreting, and synthesizing our information in the past, we’re now taking on a much greater synthesis role before we can engage in the other activities. We lack the authors that formerly acquired, designed, and assembled the knowledge message.

The Onus of Labour

The increase in shallow atoms proportionally requires that we spend more of our knowledge acquisition activities in designing and assembling the knowledge we seek prior to considering it as a whole. We must do more of our own authoring before analyzing and understanding. Previously, with a deep unit, we could accept the assembled whole (I refered to this as the author’s message in my last post), consider it, analyze it, and synthesize our understanding together. The big difference is that analyzing a unit is quite different than analyzing atoms. Synthesizing the knowledge of the unit feels very different from synthesizing that of the atoms, which have to be selected first. I think it is for this reason that search engines must work much harder to help us answer the second question if the companies operating them wish to continue dominating our knowledge acquisition activities.

Consider many of the new technologies being offered on the market. Gadgets like GPS devices do not simply enable mapping and direction giving knowledge, they’re increasingly used within the context of revealing restaurants, stores, etc. I’ve read about technologies being tested to visually superimpose meta information on objects of everyday life: ingredients on food in stores for example. In my own neighbourhood, there were a number of old houses on some streets that had plaques identifying them in a special way that let people know they could dial a number on their mobile phone as they passed to hear historical information about that area. These technologies offering context-relevent atoms, seem to exist in recognition of the need for assistance in assembling knowledge from the quantity of shallow atoms flooding us.

I’m not convinced that what is called “Web 2.0″ advanced our interaction within our world much (it just formalized some of it better). However computer-assisted shallow atom assembly could significantly advance our interaction through substantial knowledge acquisition, on top of the massive problem search engines created on the shoulders of the Internet.

Aside from basic needs satisfied from an atom or two, will search engines cease being simply search engines and instead truly assist users in substantial knowledge acquisition needs? They’re not doing this now but the seeds are already being planted (intentionally or not). Google has seen the need to insert related context links in much of the results it presents and is collecting more intelligent sorts of feedback from its users. Steps in the right direction, sure. Additionally, Google’s push of its OS to mobile devices says something about the need for a search engine to be constantly assisting our lives. Our handy pocket search. But that had better become something that assists me with the multitude of shallow atoms rather than just present me with more of them in an increasing quantity of contextual situations. Facebook with its multitude of information streams (or other current social networks) are early, lucky, symptoms of the desire for what is needed ahead.

A crude example is that search engines could capture the amount of time someone spends reading a page of search results. They could capture the time over the course of multiple “next, next, next” button pushes in a set of results. Each search could be a category within a hierarchy of a knowledge narrative. The search result links prioritized and selected from (for example) Google’s algorithms for page rank and relevence. After several sequential searches, or searches with related topics, the engine could display a page of introduction, body content, and conclusion, that provides a comprehensive knowledge narrative the seeker could acquire.

That system might present a body of shallow atoms in a wholistic message as if by an intelligent author, yet it results as a sympathetic mirror of the user/seeker’s activities. But as I said, that’s a crude example-I’ve neglected many problems and details with how something like that could work. I’m sure there are much more sophisticated things that could be (and must be) done to develop computer-assisted shallow atom assembly.

[side note: after writing this last week, I came across an article about an exciting new application, Wolfram|Alpha that may be a step in the direction I've been describing. Got to go try it out]