The average worker of 2002 has more technology at their fingertips than many entire organisations would have commanded in the early 1980s
Twenty years ago, enterprise IT was almost non-existent. Typewriters were still king; there was a fax machine in the mailroom, and perhaps a smattering of eight-bit CP/M or Apple computers around the place. You may have bought some IBM PCs, but they ran a version of DOS that was to all intents and purposes identical to CP/M; the first PC clone, from Compaq, wouldn't turn up until 1983. Bigger companies had mainframes and minis, but they ran specialist software and terminals were far from ubiquitous. Email, mobile phones, windows and mice, even the workplace LAN were still in the future.
The average worker of 2002 has more technology at their fingertips than many entire organisations would have commanded in the early 1980s -- desktop and mobile phone, PDA, PC with high-end graphics card, LCD monitor, all with one or more processors, most of which run at least fifty times faster than that in the original PC. The desktop computer's own processor has a clock speed getting on for a thousand times faster, and can do around eight times as much per tick of that clock: it still spends just as long doing nothing waiting for the user to press a key, though. It just waits much faster.
Enterprise's biggest revolution in IT has been communications, not computing. If it's hard to remember that the Internet has only been a significant influence on enterprise IT for the last three or four years, it's harder to foresee the next twenty years. One of the biggest and most far-reaching changes will be in location-based services, where a combination of ubiquitous very fast radio networks and radio-frequency ID (RFID) tags will mean that physical items -- and people -- will be as well integrated with the Net as web pages are now. Stock control and tracking, even past the point of purchase, will be as simple as searching a database. The technology exists now: the legal and regulatory implications are still to be thrashed out. People are already used to the idea that the office can be in touch via mobiles, and are just realising that a combination of wireless PDAs and VPN access to the work LAN means that their work can follow them around just as easily: this idea will expand until your virtual presence in the office will be possible, even expected, no matter what you're doing or where you're doing it.
Some other traditional 'office of the future' ideas will at last become common. As workplaces become stuffed with more wirelessly networked devices -- including fabric components like lighting, heating, security and fire sensors -- and voice recognition finally gets good rather than acceptable, you'll be able to ask questions of your systems wherever you are. Nobody's expecting the keyboard and screen to go away: as a focus for thinking and creating, reading and writing are core skills unlikely to be supplanted by technology. Paper seems unlikely to become an endangered species any time soon: the desktop of the future will probably have a large, cordless screen, a cordless keyboard and a mouse, buried somewhere under reams of printout. But you'll be able ask out loud "Where's my mouse gone?" and it'll bleep helpfully.
Security is one of the biggest unknowns. It's already possible to identify someone on camera by a number of ways, and it seems likely that should a system want to check your authorisation it will already know who you are and where you are -- passwords, PINs and explicit security devices will be no more needed to access your data than they are needed today to let you join in a conversation with your friends in the pub.
A dystopian view of future data security will have one or more quasi-governmental organisations monitoring all data usage for copyright, licensing and patent infringements, at all times, backed up by comprehensive logging and microbilling. If owning and using a car is one of today's best ways to run up bills and break the law, making use of other people's data may overtake it in twenty years' time. A utopian view has wholesale reforms of intellectual property laws, fair use enshrined in law and anti-monopolistic safeguards in place: not only will you be able to inspect your company's own performance in any area at any time, but you'll be able to put it in the context of the markets, your competitors and what-if modelling, using comprehensive demographic and historical data.
This will be another aspect of future computing that seems likely to happen, and to make a huge difference to everyday enterprise computing. Distributed processing, storage and broadband networking means that data anywhere will be accessible, and local data analysis will contextualise and summarise what it actually means without it having to be brought to one central database and crunched there. A lot of artificial intelligence research, currently of academic interest, will mean data doesn't just sit there as patterns of bits on a disk, but will carry with it a whole skein of links to related concepts. Who runs this, and who owns the rights to use the contexts, will be more important commercially than who wrote the software in the first place. If anyone did write the software: large projects are already limited by our ability to model, manage and test the processes of design and implementation, and tools to automate this will become as important to software production as they are now to chip design.
One thing remains constant, though. No matter how much processing power, data storage, new interfaces or ubiquitous networking accrues around us in the workplace, or how far the workplace dissipates into the whole world, most of the time the technology will sit there waiting for us to decide what to tell it to do. Let's hope it doesn't get bored.