Large research organizations use models to codify their findings in something approaching a consistent way. The downside is that they sometimes seemingly try to pound square pegs into a round hole--which is to say applying a standard methodology to things that are so different that they're hard to compare in a standardized way.
Gartner's "Hype Cycle" is an example. The underlying concept is pretty simple. New technologies enter the market, become the subject of breathless hype, fail to immediately live up to the most breathless hyperbole, start to therefore be perceived as failures, and finally become useful for the mainstream market in a realistic, measured way.
If you've been around the technology industry for a while, you'll recognize this life cycle in a lot of technologies. Take the InfiniBand system interconnect for example. It drove lots of start-up activity. Most flamed out. InfiniBand was frequently taken for dead. Today, it's commonplace in many of the biggest high performance computing clusters. Not pervasive certainly but not dead either. The Internet as a whole is not a bad example either.
That said, lots of technologies don't fit this life cycle neatly. Some, like virtualization, explode onto the scene and never really look back. Some just quietly become part of the fabric of everyday life. Many, of course, are the subject of much excited press coverage and never go on to amount to anything at all.
Gartner's 2010 Hype Cycle for Emerging Technologies examines the maturity of 1,800 technologies and trends in 75 technology, topic, and industry area. Many of these are, of course, pretty esoteric, but a few struck my eye as worth some broader discussion.
Gesture recognition is said to be heading into the "trough of disillusionment" with mainstream adoption two to five years out. Work has been going on with gesture input for computers for a decade of more at places like the MIT Media Lab. But with the Nintendo Wii controller it finally went mainstream and was seemingly a precursor to many more such devices.
That hasn't happened yet although, with the motion-capturing Kinect for Xbox poised to arrive before Christmas, that seems about to change. Gartner's assessment suggests that it doesn't think Kinect is going to be a transformative technology for gaming and that gesture recognition will need a few more years to work out the kinks. This isn't a bad bet, though we'll have to wait for the Kinect (and the Playstation Move controller)--and, critically, the software that uses them--to see if they hit the next level or need some replay first.
I'm less convinced that media tablets are about to peak. Tablets have been hyped aplenty of course--none more than the iPad. But the iPad has largely lived up to its high expectations. Moreover, with a much broader set of products starting to come onto the market--many powered by the Google Android operating system--this seems a category that is about to heat up if anything. Furthermore, we've barely seen the business models for content on these devices begin to play out.
To be sure, many of these coming devices and plays for monetizing content will fail. And vendors and content providers will doubtless be disillusioned as a result. But the news for users is mostly great.
Brooke Crothers of the CNET Blog Network also took a look at the Hype Cycle in the context of the iPad, 3D, and augmented reality.
Gartner views cloud computing as at the peak of inflated expectations. In this context, I assume Gartner is talking about public clouds given that it lists private cloud computing as a separate technology that has a way to go on an upward trajectory. I can't really disagree with these assessments.
It's become clear that public clouds, especially in the context of the mega-service-providers, have a big role to play in the evolution of computing. But it's also become clear that there is no "big switch" in the sense that everything is going to move to these providers.
As for private clouds, adoption in still in the early stages. I see backlash here and there about widespread "cloudwashing" but I see nothing but interest in layering automation and management on top of virtualization--whether this goes by private cloud or some other name.
Perhaps the most valuable thing about the Hype Cycle model is to remind us that fevered expectation often leads to disappointment. (The biggest forecasting trick is spotting the exceptions.) But that just because something is hyped, doesn't mean it won't be real and useful someday.