Beyond addressing the general linguistic challenges (not simple), the first step in getting there is in the disambiguation of the question being asked. That may be possible from a single more complete input - like “when was Barack Obama born?”. It may also be derived from the historical context of questions asked. For example, if the next question entered was “and Hillary Clinton?”, it would depend on the previous, unambiguous question for it to be understood that you were asking for her birth date as well. Questions can also depend on common topical knowledge or current events. Questions like “What is kobe’s percentage from the line?” would require knowledge of basketball to understand, while “What happened to davidson?” would require knowledge of what is currently happening with the NCAA tournament.

Doing this accurately - outside of more narrow domains - can be difficult.

Once we can understand what specifically is being asked for, the second step in this type of search experience will be defining ways to ascertain the “accuracy” of specific possible response sources. This may happen via a reputational model that measures historical accuracy, or a “wisdom of the crowds” model that gives you the most cited response. Both have merits.

The third element, and one that is often overlooked when this type of interaction is envisioned, is what type of commercial model can be built around this type service. Would a “meta-google” have the right to mine everyone’s tags and create a massive database of knowledge based on this? The current keyword model “kind-of-works” because a search generates links to pages - not any actual answers. If those page links went away, the ability for sites to commercialize their efforts (using the current model) would be reduced, and we would likely find ourselves embroiled in years of nasty copyright litigation that would cripple innovation in this area. There will need to be some business innovation here to make this work for all of the involved parties.

With that as a consideration, my belief is that we will see a hybrid model develop in this area that will still return links, but will deliver you to a marked-up page with the most relevant parts highlighted. Given the more accurate nature of this search model, the returned links would probably be fairly well focused and serve as a good starting point for “random discovery” type browsing. That said, the concept of “Page Rank” will need to change since relevance is far more deterministic in this search model. That’s an important area I’ll need to give more thought to.

Thanks for reading and commenting!

-john

If you keep the view that what you see today on Facebook or MySpace is the extent of the Social Web, I would agree with you that it really isn’t a significant development beyond a small, young demographic - and is more a fashion than a foundation.

But the “Social Web” is actually more than that, with applications that go beyond the ‘vanity pages’ the media tends to focus on.

Google’s search heuristics are anchored in the social web, basing relevance on the preponderance and quality of links to specific sites. Recommendation systems like those found on Amazon or Netflix are also social in nature and depend on a community to derive value. This is also true of music recommendation sites like Last.FM or Pandora. There are also emergent segments like creating topical bookmarks in browsers that are populated based on what other “peer” browsers look at and bookmark.

Many of the most interesting uses of the social web today don’t require people to ever interact directly - they simply provide increased value to everyone based on the self-interested actions of members of the community. And these are services that reach across a broad demographic of web users - young and old.

There are also significant opportunities for social technologies in the professional markets. The financial space, an area that I have worked in for quite a while, is essentially one large social network - and has been since well before the internet or ARPANET even existed. It operates based on many components that are struggling to make their way into the online world - trust, reputation, security, permissioning, and semantically driven context. These are elements that are applicable to most other professional domains that require interaction, and are all foundational social capabilities that could find their way into the foundation of much of what Apple produces.

There is so much more that can be done in the social web space , and I think that Apple is very well positioned to contribute.

As for my criticism of Apple…

If you get a chance to read back through my posts on Apple, I think you’ll see that I have been very generous in my praise for what they have done. Whenever I criticize, I do it through the lens of a loyal Apple user prodding for improvements in things I see lacking. I am a major fan of Apple’s products, and have based a great deal of my production workflow on Apple’s hardware and software.

However, the latent problem for many companies that achieve great success is that it can mask many underlying weaknesses - weaknesses that can become far more visible and significant in more challenging or competitive markets. To frame this historically, the seeds of Microsoft’s current problems were sown at a time when their stock was soaring and they were THE unstoppable force in the computer industry. There is a Darwinian nature to the technology sector that punishes the complacent and pushes for change - and that force tends to create opportunities for new entrants.

While I in no way mean to compare Apple and Microsoft as companies or cultures, I do want Apple to avoid a similar trajectory to Microsoft in the market. Nothing would make me happier than to see Job’s & Co. “laughing all the way to the bank” for a long time into the future!

Thanks for reading the post and taking the time to comment!

-john

The “social Web”– gag me. I know there are people out there addicted to the social Web; however, I have to wonder how many of those people are past puberty. More specifically, how much disposable income to these people have?

For me, and probably millions of others, the best part of the Web is that I don’t HAVE to have a lot of the kinds of unpleasant social interaction that I was forced to have prior to the Web.

First, Vista is simply not a good product. They botched the job. It doesn’t matter what it could have been, it matters what it is.

Second, I think PC operating systems are not going to go away. Ever. Somebody will produce them. We will use them. Right now, 90%+ use windows. This means the hardware and the software that we want is for Windows. Unless Microsoft do a very very bad job going forward, people will not switch to another incompatible operating system (e.g. Apple).

They key to a shift in the marketplace is compatibility. IBM compatible PCs were the breakthrough in the 1980s. If somebody came up with a Microsoft compatible operating system, Microsoft would be in real trouble. We don’t need anything different, anything fancy, anything much better, we just need it to work.

What I would want is the equivalent of Foxit (programme 5% of the size of Adobe Reader that does 90% of the job in 1% of the time) in the operating system world. It should be Windows compatible, 10% size, 10x speed, lower cost, possible free.

Please?!

just came across the blog. good stuff!

Dave

We’re working on this problem with specific attention to investment analysis and are looking forward to thoughts on equity research and to your series in general.

Regards,

Mark Soper
Alluvial Labs

news…