So we have a small problem in Australia today, and right now, it's becoming a thorn in my side!
The trajectory of the web experience at the moment is moving very much towards turning up the signal and filtering out the noise. The only way this is going to happen is by incorporating the "human" element into mashups! (Like doing away with broad spectrum RSS and having learning RSS readers that choose relevant content from the web based on my "current" interest profile based on markers that match with other publishers)
But before we can do this, we need a platform in Australia to host the next generation of mashup technologies, and it's not just the map style services, but it's providing a publishing model that allows me to create a simply mashup provider service using my favorite IDE and publish it straight to an online marketplace, locally, and immediately start making money!
So what is getting me excited, well off course, it's an idea...
A while back, Yil and I were looking for a cool cafe to hang out at for the arvo while we were in the city. Now, as most people know, there is nothing like human referral, or as the suits call it, Word Of Mouth! Now, we stumbled around, found some place, had a cuppa, and eh, it wasn't that bad.
That night I got home and thought, well, what I would have loved to do is fire up a small app (web or mobile, don't care) on my smart phone, punch in the words, "Good cafe near the State Library" and voilà, a little profile comes back with a couple of choices. It would be easy, the app would follow the path below:
1. Check location of phone <- Map Location 2. Check keywords <- Analysis service returns keyword matrix 3. Parse Matrix <- "Good" marked as emotion/rank/etc, "cafe" marked as shop/service/etc, "near" marked as location/proximity/etc, "State Library" marked as location/waypoint/landmark. 4. Resolve Matrix <- search web for blog entries, forum comments, any live commentary/opinion/etc for elements of matrix 5. Present Profiles <- rationalize resolutions into actual shops/services
1. Check location of phone <- Map Location
2. Check keywords <- Analysis service returns keyword matrix
3. Parse Matrix <- "Good" marked as emotion/rank/etc, "cafe" marked as shop/service/etc, "near" marked as location/proximity/etc, "State Library" marked as location/waypoint/landmark.
4. Resolve Matrix <- search web for blog entries, forum comments, any live commentary/opinion/etc for elements of matrix
5. Present Profiles <- rationalize resolutions into actual shops/services
The point is, the next generation of solutions will be wired together by taking the most innovative, high-value services from the web and orchestrating them in a way that provides true value and innovation to end-users. And the barriers to innovation and invention will be lowered because the workload is being spread across the whole landscape, not just within companies who can assemble very clever people. And also the concept of integration will become a problem past, just like the ubiquity enjoyed by technologies like browsing, email and SMS today.
And where do the choices come from...well, that's easy, the Aussie Web Platform (and some mashup muscle). So if the first issue is the publishing platform, the second is the variety of services, because maps only get me so far. So I need some categories of services that include some intelligent services (like one that scans blogs for keywords including locations, identifiers like shop names, and emotions to gauge suitability), some object meta-data services (address, opening hours, byo, etc), some presentation services (mobile platform, layout, style) and so on.
Then the scene is set for the local software economy (the suits have a word for this too, LSE) to consume these services, make some cool businesses upon new offerings, and money floweth through OzLand!
Why do we need this? Because this is the path humans are taking. And like everyone else in the world, the distance between us and them is now measured in Mbits! And we can continue to force our latest generation of web innovators to seek housing and support OS, or we can build the infrastructure to support it locally. And this local footprint is even more important when trying to coax the big players who are keeping their high-value service platforms locked inside big datacentres in their basement to publish out to the world. The risk to hoisting those systems into the cloud elsewhere in the world is a lot higher than if it were being done my a local Aussie player on infrastructure located within our shores!
And trust me, while I've already heard so many reasons why people think banks, government departments, and other LIOHWSI (large institutional organizations hiding wicked systems internally ;)) would not hoist their systems into the cloud, what I'm really keen to hear is ideas on how they might?
How say you Australia!? Ours or theirs!?