Dan Farber of ZD Net reports and comments on 10 predictions to shake your world, based on a recent GartnerGroup report. You can read a whole slew of articles about these issues in a special report on the Gartner Symposium/ITxpo 2002.
Without deconstruction the article too much, here are my comments on Dan's comments about Gartner's predictions:
1.) Bandwidth will become more cost effective than computing. Dan says it'll take a decade before you see companies "pry the big hard disks out of people's computers." I think we will never see the end of local storage, that it will become so massive that we will hardly think about the problem of storage. Local storage, which will likely be in your wall somewhere, behind a panel like today's circuit breakers, will consist of Terabytes of disk storage and Gigabytes of random access memory, along with multiple processors, linked to devices throughout the home or office via wireless to conduct real-time grid-computing like functions. This will allow "dumber" (even though they'll be a lot smarter than today) clients to perform more complex functions.
2.) Most major applications will be interprise. I'm with Dan that it will take at least half a decade to see much progress in this area. Gartner analyst Carl Claunch also points out that there are a lot of political reasons the tools we use will never work together. I'd add one more reason: Intra- and interdepartmental politics and inter-enterprise demands for secrecy and an "edge" over even your closest customers will always be a barrier to successful implementation of comprehensive interenterprise applications.
3.) Macroeconomic boost from interenterprise systems. Dan's a skeptic, but ultimately believes Gartner's prediction that this productivity gain will come in the next ten years. I think that even if the systems problem is worked out, the human political problem cited above will dampen productivity gains. It takes a very long time for people to change. That's not to say there will not be continued productivity gains, but really huge gains are unlikely.
4.) Successful firms in strong economy lay off millions. Dan remains a skeptic. Gartner believes companies will be letting millions of workers go due to automation. There are two reasons not to expect this in the next decade: First, capital expenses for new automation have to be justified by economic conditions -- we're in a slowdown now and continuing layoffs will only lengthen that slowdown, reducing the amount consumers spend, which drives capex. Second, automation often creates additional opportunities for human work that did not exist before. Once you take the filing out of the hands of people, you begin to need people to analyze what the automated filing is telling you. Where the white collar laborer (that is, the non-executive) is replaced by computational capabilities you often need a business analyst to replace them. This means that automation will require massive retraining to succeed and this is not the environment or the government to make that happen.
5.) Continued consolidation of vendors in many segments. Dan says "duh," and so do I. But where there is consolidation, there is also new opportunity as portions of the value chain are disaggregated and outsourced.
6.) Moore's Law continues to hold through this decade. Dan agrees. The industry keeps waiting for Moore's Law to break down and that will happen about the same time that we experience a loss of gravity. While transistors may ultimately be only one atom wide and can be shrunk no further (who knows? they might go smaller), the geometries of chips can easily be increased for years to come to accommodate more compute capacity.
7.) Banks become the primary provider of presence services by 2007. Dan doubts standards can be worked out by 2007. I wonder whether there will not be a whole new layer of economic activity aimed at managing our privacy, preferences and, in some cases, anonymity. Banks are one natural provider of these services, but I don't see preferences management as a core competency in those organizations. Something else will happen, though it is certain banks will be involved somehow, since they deal with money.
8.) Business activity monitoring is mainstream by 2007. Dan doesn't really say he agrees or disagrees with this prediction, just points out that it is closely related to prediction #2, which he says will not come to fruition in this decade. I think that the idea of a few people running most companies through largely automated systems is silly. We'll track many more activities, but we'll be employing people, and people slow decisions down. That's a good thing. An economy running on automatic would be a disaster waiting to happen, especially given the willingness of terrorists or even just competitors wanting to bring things to a screeching halt. Business activity monitoring is a buzzword that will go away.
9.) Business units, not IT, will make most application decisions. Again, Dan reports this prediction without really judging it. I think it's a given that IT decisions will devolve onto the people that will use the technology, but that does not mean that the IT organization will vanish as much as it will melt into working teams. This closer linkage between IT professionals and the people they support will improve specifications and usability.
10.) Pendulum swings back from centralized to decentralized. This follows from #9, but will be restricted by the evolution of bandwidth to and from the home. It's my contention that the Digital Cottage, a home with adequate compute power and sufficient bandwidth to participate in enterprise applications (both up and downstream) will transform the economy from one where people go to places where they work together for a fixed number of hours a day to one in which people work together throughout the day or night, on their own schedules (and against clearly defined goals), while taking greater parts in the life of their family and community than was possible in the Bubble Years.Posted by Mitch Ratcliffe at October 11, 2002 03:27 PM | TrackBack