7 Definitions of Web 3.0

A few weeks ago, Google’s Eric Schmidt managed to sound like both a n00b and a tasteless PR hack in the same 2 minutes, while trying to define “Web 3.0.” I wanted to try and round up some useful things about “Web 3.0″ that Schmidt could have said. So - to that end, here are 7 definitions of “Web 3.0,” ranked roughly in order of plausibility, coolness, and common-sense compliance.

7 & 6. ReadWriteWeb’s Web 3.0 Definition Contest

RRW held a “Define Web 3.0″ contest to give away tickets to Tim O’Reilly’s farce of a conference, the Web 2.0 (TM) Expo. RRW’s contest chose three winners, two serious.

The first is a near uninterpretable re-hash of they typical Tim Berners-Lee “Semantic Web Spiel,” only tangled in with cult-y ATOM evangelization:

“Atom, the Atom API and semantics, particularly Micro formats initially, are the constraints that will make this happen. Atom features not because of technical merit but by virtue of it’s existing market deployment in a space that most EAI players won’t even consider a market opportunity. Hence Web based components start using Atom API as the dominate Web API – Feed remixing is indicative. Atom will supplant WS* SOA.” [from RWW]

The second can be summarized with this gem:

“Web 3.0 to me is video-driven social networking, video-driven news blogs and user created entertainment, and video-driven email (yep..one day!).” [from RWW]

OW MY BRAIN. Are you kidding? Video email is the best you could come up with for “Web 3.0?” And RWW liked this definition enough to pay you for it? Dude, why don’t you just sign up for 3D Mailbox? The future is here today!

5. Business Week’s Stephen Baker

Stephen Baker apparently lead a panel on “defining Web 3.0” in Monaco (what’s the broadband penetration there, I wonder?), almost a year ago. Unfortunately, the results they came up with can be summarized as “more 2.0.” More cheap, more pervasive, more access, better data security. Damn, Stephen, that sounds like a service pack release, not a new paradigm. Definitely worth holding that high-powered panel in a swanky location.

4. HowToSplitAnAtom’s Definition

This definition doesn’t seem to offer anything new over what we have now. RSS feeds? Search verticals? Results clustering? Search by content type? Sounds like the sort of well-intentioned but uninspired upgrade Yahoo or Ask would make to their search engine, and the sort of mathematical mojo that’s already incorporated into Google’s algorithm.

“Web 3.0 will take this one step further. If you are searching for information on Cars, for example, you would use the search engine as you normally would, but your results would be more specialized subengines. I would find BMW Search or Kia Search. From there, I would be able to dig deeper and find items that have been tagged as relating to BMW and sort them into their major categories (pictures, videos, blog posts, news articles, commerce etc…) Each of these could be captured as an RSS feed so that I can be alerted when something new is added to by search profile.” [from HowToSplitAnAtom]

3. Phil Wainewright - Layered Architecture, New Business Models

Phil tackles Web 3.0 from the enterprise direction, laying out a three level structure that he expects the next generation of companies to work in. API Providers are at the bottom, Service Aggregators act as middlemen, and Application Services (consumable products) sit on top. Profitability grows from the bottom up, with Applications being money-makers, and API’s being commoditized.

I’d like to make one thing is absolutely clear right from the outset: Web 3.0 isn’t just about shopping, entertainment and search. It’s also going to deliver a new generation of business applications that will see business computing converge on the same fundamental on-demand architecture as consumer applications. So this is not something that’s of merely passing interest to those who work in enterprise IT. It will radically change the organizations where they work and their own career paths. [from ZDNet]

Huh. Not bad - Wainewright has laid out an architecture for the delivery of next generation services, which certainly counts for something. What he doesn’t specify, however, is what those next generation services may or may not be - for the enterprise, or consumer spheres. Wainewright has a whole series of articles on this theme, however, which I admittedly haven’t had time to sponge yet.

2. Valleywag: Web 3.0 is still Web 2.0

Valleywag points to two examples of 3.0′ish companies - Radar Networks (semantic search engine), and Spock (people search engine). And then relates succinctly why they aren’t really 3.0 at all:

The problem with Spock and the greater ideal of a semantic Web is the continued need for human input. Intelligent tags for metadata don’t magically appear. Humans need to establish the relationships between data points. Even Spock claims to rely on users to ensure personal data is correct. Spock is going to have a hard time keeping tabs of all 6 billion people. How exactly is a semantic Web going to manage a world’s worth of data? Wikipedia can barely keep itself straight. [from Valleywag]

Well said. I love VW. These 3.0 services are just aggregating 2.0 data and functionality - an incremental step, but hardly a leap. As such, VW’s hasn’t so much provided a definition of 3.0 as it has eliminated pure aggregators from that definition. That’s an elimination I agree with; aggregation can add value, but crafting a paradigm shift out of remixed existing data is a big step.

1. Tim Berners-Lee and the Semantic Web

Yup: #1 is still the seminal definition, from the father of the internet. In a nutshell, TBL’s 3.0 is much more structured markup and automation via a device independent, hyper-connected services architecture.

In this version of the Web, sites, links, media and databases are “smarter” and able to automatically convey more meaning than those of today. For example, Berners-Lee said, a Web site that announces a conference would also contain programming with a lot of related information embedded within it.

A user could click on a link and immediately transfer the time and date of the conference to his or her electronic calendar. The location - address, latitude, longitude, perhaps even altitude - could be sent to his or her GPS device, and the names and biographies of others invited could be sent to an instant messenger list. In other words, the “mark-up” language behind each Web page would be cross-referenced into countless other databases, once developers agreed on a common set of definitions. [from IHT]

My thoughts? The interactions and automation enabled by pervasive, well-structured and mappable meta-data would truly herald a “next generation internet.” But - at the core, this vision seems to be incomplete. While the means to store and communicate meta-data may be under construction as we speak (i.e.: evolving standards), the means of automating the creation of that meta-data is less so; without the always-in-the-future “intelligent agents” working to categorize & connect data, we fall back to algorithms, keywords, and the fragmented universe of human-powered tagging - which is where we are already.

So - while I can appreciate the vision and see its utility, my gut feeling is that TBL is talking about Web 5.0. The notion of “semantic information” has much thinking to go into it yet before it will truly revolutionize our day-to-day lives.

In Summary

The variance in definitions makes clear that 3.0 is still a moving target. The definitions out there oscillate between minor increments to the functionality paradigm we enjoy in 2.0 to AI-enabled visions of a connected future (5.0?). The reality will probably be both stranger and more boring than expected, landing somewhere in the middle.

Automated connections between databases, for example. Speaking to Valleywag’s concerns (or HowToSplitAnAtom’s vision), I think there’s still plenty of value to be found creating automated connections between human-populated databases. Del.icio.us tags could be adding semantic information to Yahoo!’s search algorithm, for instance (why they aren’t doing this already, I don’t know). Its a step towards TBL’s vision, but only about 1% of the way there.

Would integrating del.icio.us into their search algorithm make Yahoo! Search a 3.0 product? Ultimately, does it matter? As with 2.0, I think the definition will be emergent, defined gradually by a million small decisions, changes, and startup launches that take place daily. A few years from now, we’ll look back at today and be able to elucidate how whatever it turned out to be started; at the moment, however, were too deep in Weeds 2.0.

All of which makes me all the more annoyed by Eric Schmidt’s speech. With everything there is to talk about, and with all the mind equity at Google, he could have said something meaningful instead of just reading off a press-release about Google Apps.

This entry was posted on Tuesday, August 28th, 2007 and is filed under Emerging Tech. You can follow any responses to this entry through the RSS 2.0 feed.