when you start versioning things, the next step is asking, "What is the next version going to be?" Right now, there is no clear road map of what the next "big thing" in the web will be. We are seeing rumblings as Wikipedia suggest, but these are just non-specific trends being spotted. We can't even put a timeline on web 3.0 because we don't know when web 2.0 began and ended. Heck, I can't even get a good definition of what web 2.0 is since you see it as being a marketing buzzword as well as an amorphous collection of things that sprouted up as being the supposed latest trends in technology for the web (i mean, some people call iPod web 2.0. But iPod is a device, so what makes it web 2.0????)
The only reason why we're asking what is web 3.0 is the strict purpose of predicting or even knowing the future (from a business point, from a technology point, etc.). That said, we have to look at current trends that define what web 2.0 is:
- blogs
- community sites/social networking
- rss/ATOM feeds
- podcasts
- tags/tagsonomy
- AJAX
- Flash
- SOA/web services
- Mashups
- torrents
These are the overarching trends. But we need to examine things closer to see what's going on. First, I see a general movement towards the socialization of the web. You are empowering people to generate data (not just news). Now, we have tools that are more effective in taking large pods of data, aggregating it, and then redistributing it. Also, the data is becoming more meaningful (tagsonomy), allowing the improvement of finding what we're looking for (via searches).
Also, interaction has become huge in the newer versions of the web. AJAX and Flash upgrade the static feeling of the older forms of the web. Rather than having to wait for a horribly slow gif to animate or a page to be refreshed using the meta-refresh tag, we can simply use Flash and AJAX as the means for displaying content in more real time.
However, that advantage has come with the fact that more people are or have upgraded to broadband. There's a good note Reed Hastings in the wikipedia entry that talks about the improvement of broadband. Maybe users back in the day of dial up were frustrated with the slow response time, the lack of desirable content, and overpromised, undelivered technologies that emerged at that time. Now, though there are some huge advantages, partly due to the improvement of download speeds, the capacity of computers and the upgrades to the web standards.
In summary, you have four real trends: social computing, meaningful data, interaction and performance.
Those four elements are generally what I've seen in computing as a desired medium. So how do we move forward from here?
The thing about a previous generation is that a lot of groundwork will be laid down. It's the foundation for which the next few generations will build from. Here, the foundation from web 2.0 are the data, the services, the interactivity, the APIs, the increased number of users, the better understood infrastructure, the standards coming into place, and the performance.
I think a massive strategy will be improving on these areas. The wikipedia article mentions a few trends:
- AI
- The 3D web
- The web as a database
I think there are some good legs here to start. But you have to be specific in defining these areas and seeing how these will play a part.
First, AI
If web 2.0 was the foundation for teaching computers (i.e. creating monstrous repositories of data like wikipedia, google, Amazon, IMDB, etc.), then you have to figure out ways of making that data useful. Most things I'm seeing are analytical tools, like data mining, business intelligence, report generation, data searching. However, I feel that these areas are too vague and need to be narrowed for areas to be more useful. You need to ask very specific questions in building up AI services. In general, you need to ask yourself, "What is it that I really want to know?"
Here are some sample service ideas I have for web 3.0:
- I want to be able to find any image based on properties of the image. Let's say I take a picture of a building, an apartment. From that image, I should be able to derive various properties (e.g. color, architecture, date of construction, number of residents, location, material used for construction). Now, those properties ought to be made searchable. More than searchable though, I want to COMPARE those properties with another building. Why? Perhaps, I'm looking for a new spot to stay in and I like where I live, but do not like the location. Or maybe for a dating service I upload a picture of a famous Japanese singer I want to marry. The service would then attempt to locate someone wtih similar facial features in the database using a comparison of facial structures. Because of that, the probability as a customer for me wanting to purchase the service is much higher than having me search through every little picture and wasting 30 hours of my time before getting frustrated and leaving.
- I want to be able to plan my vacations better, based on my money in my account, weather conditions, location of desired place to travel, the timing of my projects and how much debt I have as well as the best price for a given ticket. I no longer can simply plan a date for purchasing a ticket. These days, there are more factors that prevent me (or perhaps many people) from executing a vacation plan. Worse yet, most people don't want to merely search week, after week for tickets just based on price. And the ticket should be purchaseable in any language as a service.
- Context keyword searches. It seems as though most searches use some form of probability and geographic proximity to extract results from a search. Imagine the time saved if we could contextualize how a keyword or phrase is located. I know there's a search engine that is more for intentions. That's partly what I want because I want something that knows the reason behind my search. That part is context. It's the difference between your English junior high school teacher telling you that Shakespeare was a great author and your college professor explaining Shakespeare's role as a Renaissance man and providing you the background of what made Shakespeare significant in literary history. Part of me believes that we need to teach computers more about a web result here. Imagine typing in "shiho" and getting a blended mix of anime characters, women whose name are shiho, and the model herself. I only care about the model herself, but the engine would only know that if we added the keyword "model". I mean, not all pages would go into depth and say, "This is Shiho Yanno" allowing you to perfect your search, since Shiho only provides her first name in Romanji rather than kanji, making your search more excruciating. I ought to be able to add some sort of filter that marks those pages I never wanted for this particular type of search as being related to some random person with the name of Shiho or an anime. Or perhaps I need to be able to mark for the search engine that those results only pertain to the model Shiho and that's the only search I'd ever want again. This all pertains to contextualization.
The 3D Web
This has always been something discussed. I remember the days of VRML thinking, "Wow! A real 3-d shopping mall!" But that idea never manifested due to the extreme complexity in building all the graphics and math for generating such a thing (not to mention the horrible slowness at the time). But I don't think anyone will forget Tom Cruise in Minority Report scanning through a holographic image. Definitely seems better for your wrist than a mouse and keyboard!
Despite this and the use of projectors, we're still really off from having 3D anything. I saw a piece of technology for creating actual objects in 3D space, but mostly for prototyping. It's not nanotechnology that has mini machines assembling packages in some microwave as William Gibson had described in one of his novels.
So we still have to be content with a limited web. Graphically, people pointed out Second Life. I think people should not limit the notion of 3D just in terms of graphics. I keep going back to memories of wearing funky blue and red sunglasses in a movie theatre for a stupid Elvira show. Not very interesting.
The 3D web for me means information should be 3D. Actually, I don't even want to think of 3D. Call it N-D, where N is a finite number and D is the number of properties a piece of information has. And that's the real catch. Making information, whether it's an image, a video, a song, or a word, contain references about it. It's like Wikipedia, except deeper. Let's say you start with a word, Toyota Prius. Now, you want to pull out all immediate references (call it 1st level references). So you might get properties of "car", "Toyota", "hybrid," "sedan," "fuel efficient," etc. 2nd level references might be "55-60 mph/hour", "four wheel," "Japanese manufactured," "environmental." And you keep drilling down or expanding your universe of terms for that data element. In technical terms, this might be an object containing an array where each increasing node represents less relevant data. And the data in each node represents a hash containing the related data and a link to the next term in line.
The Web as a Database
I have no doubt that the web is already a database. It's just horribly disorganized with too many services that overlap, making the web more confusing. I think the problem is not so much in that the web is disorganized, it's that no proscribed methodology for organizing it (outside of Google) has been thought about. The web is just a bunch of references to each other with the possibility of circular references and no true sense of data integrity. Worse yet, you get these so-called services that make the web even worse off than it already was (think of myspace and all the horrible layout tools, privacy conditions, spam accounts and whatnot).
What is needed is something to proscribe how the web ought to be. At least from a database point of view, if that indeed is how we intend to use it.
1) The web ought to be partitioned into highly relevant to irrelevant content. So there will be degrees of highly relevant, excellent, trustworthy content. Take Amazon, imdb, Netflix, cnet, etc. as examples of where you are getting top content. Then take some spammer who sets up a false domain with a fake information search page. When i hit a page, I ought to be able to return results that I want.
2) Not all people really should be able to contribute to the web. There I said it. I am prejudice. But my prejudice is derived from the amount of trash on the web. Think of the useless flame wars you see on message boards. I dislike reading about how one person is required to have an opinion because he has oxygen in his lungs but only spews out hateful, one liners that add little social value. I don't think that the web ought to be naturally filtered through browsers or content filters just based on 4-letter words (movies contain these as well as lyrics, so you're signifying your own demise based on this). But people ought to earn their spot in creating content that gets ranked based on their global net ID. Sounds a bit Orwellian, but this is a case where the social web actually loses out based on the maturity of some audience (naturally I'd get pounded too but I am attempting to write better content :p)
3) Proscribe methodologies that are consistent for website production. We've seen the MVC methodology rise up, with even frameworks such as Ruby on Rails, PHP Cake, Spring, Struts, etc. taking charge. The infrastructure needs to be similar too to prevent things like security holes from occurring. Again refer to my article about Yahoo using their knowledge to become an internet platform. Having similar structure/methodologies for websites would help improve the quality of the web as a database since sites can be more easily researched.
4) Increase the quality of information on the web. Part of this is the responsibility of the community of the web as a whole to ensure that information is top quality (e.g. wikipedia). However, we need to also improve the ways we drill down into information. As I mentioned in my recommendation for information as a data structure, we need elements like this to allow searches to be improved and provide people with the information they want. Imagine being a scientist attempting to find a formula. If that scientist had to sift through less information but information that was higher in value, he would be able to obtain his results much faster. I feel that though the results for the web have improved, I still spend far too much time sifting through results in attempting to ascertain the true answers I want.
5) Have dedicated repositories for information. I hate the "Me2" web. Okay, I admit I've done it as well because I want money. Who doesn't? But what is irritating me is seeing the multitudes of copycat websites coming from large providers like Yahoo, MSN, Google (even!!!!) that add little value. Take for instance, Yahoo 360, MSN's Spaces, or Google's orkut. None of them provide any additional functionality above what Myspace, Friendster, etc. do. Also, why is Myspace trying to become a portal site? They're just overlapping with Yahoo in Rupert Murdoch's silly scheme of making his own web portal.
These redundancies are simply a waste of human effort. I think these companies (and people in general) need focused services and products to prove to be effective. I know competition is a great thing, but it's overkill now and it's getting boring and time consuming. I want to see specific data repositories spring up. For instance, I want to be able to find all instances of data relating to Hibernate when something doesn't work for me. I don't want to have to search through 30+ forums and still not find the key result. Or perhaps I want to find all relevant information on Yumi Yoshimura up to what her farts smell like on a repository that is 100% dedicated to her. I don't want to see fan club with 15 members here or fan club with 45 members there. I see no purpose, just redundancy.
If the web becomes a database, we need to eliminate these redundancies somehow. That will improve the accuracy of information.
Anyway, that's the rant for today. Until next time, my guess is that Yumi's farts smell like chili.
Sphere: Related Content
Trackbacks: (Trackback URL)