By 2010, your computer will be making appointments for you, calling your friends to schedule dinner, and predicting your next desire. At least, that’s what my latest issue of Business 2.0 is predicting about SkyNet er, Web 3.0, aka the Semantic Web:
"In the semantic Web, your software agent will "know" in advance what’s involved in arranging a dinner. Instead of you sending out a flurry of e-mails, the agent could cull the conference attendees and make a list of potential invitees. It might also look through your address book to see which of your friends live in the city where the conference is being held. Once a list of potential dinner guests has been approved by you, the agent would negotiate the date and time with everyone else’s agents via a calendar database, pick a restauarant from another database based on availability and your personal preferences, make the reservations, and send out directions."
The big problem is, how will all this actually happen? Later on in the article they touch on the crux of the issue by lamenting the fact that people won’t mark up data until there’s software to consume it (and compelling software at that), and there won’t be software to consume it until there’s data to be consumed (and I would imagine that’ll have to be compelling data as well). However, there are two other problems that need to be addressed, which the article neglects to touch on.
That "software agent" that does such amazing feats in the article’s scenarios doesn’t exist yet, and as of right now it’s just a twinkle in a few programmer’s eyes. These software agents will serve as our personal concierges on the 
Internet, leaping into action at a moment’s notice to fulfill our every wish. Sadly, they need to be built first, and before they can be built they have to be defined.
My take is that you can’t really remove humans from the equation–the software agents cannot act independently of their human masters. What this needs to be is a system of personal servers that can interact with each other–basically a glorified messaging system. I have my agent send your agent a request for a meeting, you tell your agent you want accept, it goes on both of our calendars, done. This article was the first time I’d heard the "agent" moniker, and I actually kind of like it. "Servers" and "message queues" conjur up images of data centers and enterprise software environments, which is exactly what needs to be avoided in order for a new system of any type to be adopted by the mass market. Hmm… iAgent… maybe it’s the next killer mobile device 
Which brings up the next issue that needs to be addressed before this utopia materializes, which is a cohesive online identity system. And by identity system I don’t mean "authentication" system, which is what I consider technologies like OpenID to be. An online identity is something you can put on your business card, something you can convey to somebody in a conversation, something that people can refer to as "you" on the Web to find you, just like your name in the physical world. It’s also a handle to your agent. People’s agents will need to communicate and find each other with the identity system, and this is not something that technologies like OpenID address. Without this capability, the system won’t be elegant or easy enough for people to adopt en masse, and the learning curve and setup time will outweigh any perceived benefits.
