The rate of change of computer technology is phenomenal. Who, seven years ago, could have predicted a PC operating system would emerge, namely Windows NT, which would rival Unix. Who would believe that an application suite such as Microsoft Office would need more than 170Mb of disk space to install.
Yet today, PCs are infinitely faster than their predecessors. Each new generation of microprocessor scales dizzier heights of performance. Supercomputing is almost within the reach of the desktop user. Computing has never been cheaper.
The new world order of power at the desktop has opened up the opportunity for innovative applications of computer technology.
We're all used to bottlenecks like disk access times, but what about the screen? A 15in or 17in monitor limits the ability of information technology to change the way we work. Imagine doing any physical activity through the frame of a screen - it's painful. Mark Bird, group marketing manager of display distributor Steljes, believes some companies can already supply the future.
"Fujitsu has just launched a 42in flat panel, six inches thick. Displays like this enable digital convergence - PC, video conferencing, broadcast video all using the same screen." Such panels cost around u9,000, but Bird expects prices to drop quickly, reaching $100 an inch in a couple of years, and fitting a PC budget within seven years.
Moving further into the future, panels covering walls or a desk become feasible. Documents will be seen full size, manipulated directly on a surface that is aware of hand movements. The liberating effect of such large screens has to be experienced to be believed.
Agents good, keyboards bad
"Agent" is an overused term, often applied to little more than macros, but intelligent support is taken very seriously by Microsoft. Jonathan Hulse Office product manager at Microsoft, says: "We've introduced the Office Assistant, an intelligent tool for just-in-time help. This is just a start. We will be building on this approach to put more intelligence on the desktop. We already have prototypes building on the natural language approach." At the moment input is typed, but soon voice will play its part. "There has been a lot of work done on voice, but it doesn't work very well yet. With natural language we've done a lot of the groundwork needed to cope with the syntax of speech," says Hulse.
Sam Sethi, marketing manager at Netscape UK, agrees: "Voice will be the next big technology, and will become the primary means of input in 10 years. Accents, and probably a new slang language, will occur as we shorten or create new words to interact with devices. Pen technology will be another common means of input. The keyboard will hopefully be dead."
Vincent Smith, communications manager at the IBM PC Company, points out that speech is more than just commands. "We already have prototypes of a handheld web explorer that respond to speech commands. But (in the future,) speech will be a conversation. Intelligent assistants will then go across the network, executing your tasks." Smith also points to developments such as head-up displays and interaction by monitoring eye movement - particularly useful when reading documents. It seems likely that the keyboard will become a specialist tool, and be eventually sent into oblivion.
Dumb PCs, smart watches
PCs still figure large in Microsoft's vision, but Quentin Stafford-Fraser, research scientist at the Olivetti & Oracle Research Laboratory, prefers high specification, relatively dumb displays fronting powerful networks.
"You won't be tied to a desk or a particular computer, the information will follow you around. Whether you are using a device on the desktop or your mobile phone or a cashpoint, your personal information and interface will come up."
Before information can play tag, you need to be recognised.
The Cambridge-based laboratory has used intelligent badges for several years. As individuals move around the site the badge tracks them. "You can go up to a workstation anywhere, press a button on the badge and your own screen pops up," says Stafford-Fraser. There are other applications too. "We have a coffee vending machine that uses the badge to make drinks to your preferences. It's a very simple interface; the badge only has two buttons, but the system knows who you are and where you are and can respond accordingly." In the future such technology could be worn as a watch or embedded under the skin. Imagine an intelligent phone. As you approach, not only does it know your dialling habits, but it adopts your telephone number for incoming calls.
Before the PC model becomes redundant, the problems of centralised computing must be overcome. Stafford-Fraser is confident they will be. "Until recently networking was the bottleneck - soon it won't be. As the novelty of personal computing wears off, most people would rather someone else did the maintenance for them. You would still have your personal environment, local peripherals and so forth, but not the big noisy machine on your desk." Network reliability has been a problem, but Stafford-Fraser sees replication making this less significant. And for those worried about returning to timesharing, when peak periods meant s-l-o-w processing, he points out there is no reason why you can't still have a dedicated processor - located in the computer centre, rather than on your desktop.
Professor Steve Molyneux of the Interactive Communication Technology Research Centre at the University of Wolverhampton, shares the vision.
"The first thing you have to do is get rid of the PC. You end up with an information appliance, a transducer between the information on one side and the user on the other. With network-oriented computers the software will reconfigure itself to deal with the requirement. You can see an analogy with the book - the shell is the same but the content changes in different contexts. "It boils down to high-speed networks and getting rid of the computer. No one wants a bloody computer," says Molyneux.
It's good to talk
Today globalisation is a big issue. Large companies handle business across the world, and language barriers eclipse European Monetary Union's problems.
Monica Beltrametti a spokeswoman for The Rank Xerox Research Centre in Grenoble which specialises in linguistics, says: "(In a global corporation) you need to retrieve documents in French, German, whatever. We're looking at access tools that can take a query in your native language, find the documents and provide a translated summary. The ultimate goal is automatic machine translation. If the document is very technical with a restricted vocabulary it's much easier than doing Shakespeare. That's almost possible today. But general documents have ambiguities that can produce difficulties.
What does a pronoun like 'they' refer to? The human brain is very good at sorting this out; but within 10 to 20 year timeframe this is perfectly achievable (by computers)." Given parallel developments in speech recognition, Beltrametti sees no reason why there could not be Star Trek-style instantaneous speech translation too.
Breaking the paper chain
While the paperless office is a long way off - paper is, after all, a superbly flexible technology - a big change looms. Rob Allen from workflow and document management specialist SNS, says: "At some point in the next 10 years there will be a huge change of mindset. I don't know if it will be gradual or sudden, but we will go from documents being paper items to electronic. Printing things out will be taking a copy of the original. This will facilitate a huge change in the way we work."
Allen sees electronic commerce taking off, with most transactions made without paper. And there's more. "Let's assume communication bandwidth and cost is not an issue. You can have virtual companies. There's no reason any longer for physical co-location; you can work anywhere just as well." To make this all work there will be greater demands on the service provider.
"A turning point will be service providers guaranteeing delivery in a timeframe. If you had guaranteed one hour delivery of a message, you could use it for a bidding process."
Make or break comms
While many of these technologies will evolve in 10 or 20 years, a big question-mark hangs over communications. Ultra fast networks will be achievable on corporate sites or in cities, but is it realistic to think that they will be available nationwide? Stafford-Fraser of the Olivetti and Oracle Research Laboratory believes so: "You can get a lot down ordinary telephone cables, it just requires the right hardware, and technologies like cable modems extend what is possible. People will do a lot more from home, and I personally hope that telecommuting will revive village life."
The decision is more political than technical. Chris Yapp, managing consultant of ICL Interactive, says: "When you try to project you get a range of scenarios with broadband in the cities and not much more than ISDN in rural areas. This could result in a reconcentration in the cities, but if environmentalism wins we might see more universal services. Think about getting jobs into places like Devon and Cornwall. By putting in broadband it would be possible to provide employment without the environmental impact of industry." Private funding will probably be necessary. "At the moment I find it difficult to see widespread networks paid for from the public purse. One possibility is to have virtual private and public networks running over the same infrastructure, like 0800, 0345 and 0898 numbers all on the same telephone network," says Yapp.
Netscape's Sam Sethi neatly describes what the technology will blow away.
"Today I travel to a central place of work. I then meet with other people internally or externally. Occasionally I travel further, involving car, train or plane. We all commute to and from work at the same time. Result: congestion, pollution and frustration. In an office we have separate Email, fax and phone systems both mobile and desktop. Information overload is becoming a serious problem. The people we spend most time with are work colleagues not the people we always choose to. Culturally we therefore cram in activities such as leisure and shopping into defined hours. " The emerging technologies will enable us to change all this to a more people-centred way of life. The opportunity is there to tame the office and put work where it belongs.
Intel: growth of processor capacity
Doctor Albert Yu, general manager of the Microprocessor products group at Intel, forecasts an incredible increase in processor capability.
The chips of 2011 could contain 1 billion transistors and handle 100 billion instructions per second. Compare an average processor of today with Yu's forecasts.
Such development has a price. Design, testing and controlling power consumption are intensely technically challenging, while the sheer cost of development, maybe $60 billion u37 billion), will be immense. Even so, Intel is confident that the obstacles will be overcome, and processor power will keep up with demand.
Thanks to this growth, Intel does not see personal parallel processing becoming a major demand. A spokesman Julian Powell, says: "There is not much that a CPU can't do. We're spending a lot of time developing the whole platform, dealing with delivering information fast around the system.
A Pentium II can do a heck of a lot now; the constraint is not the processor and we'll drive the technology to make sure that the processor stays ahead of the market." At any one time a function - MPEG, 3D graphics and so on - might be handed off to a secondary processor, but by the next generation Powell believes it should not be necessary.
Futuristic dreams: John Sculley
In the late 80s, Apple chairman John Sculley dreamed up the Knowledge Navigator. Short videos were made of individuals using a mock-up of this conceptual device. Some have been critical of Sculley's vision, like , Stephen Levy in his excellent biography of the Macintosh entitled Insanely Great. "Knowledge Navigator was a pastiche of Bush, Englebart, Kay, Atkinson and Dick Tracy. Every aspect of it had already been well traded in the marketplace of ideas; much of the enabling technology that would be required to realise the device was already being developed in laboratories waiting for Moore's law to drive down the price. Yet Sculley won acclaim for his futuristic paste-up job," wrote Levy.
This is unfair. The power of Knowledge Navigator was seeing it used.
In one scene a university professor chats with an intelligent agent handling his real-time communications. It retrieves information from a global source using fuzzy specifications, then translates the data into a beautiful graphic simulation. Finally, it links in a remote colleague for a video discussion of the simulation. Knowledge Navigator is bound to be wrong in practically every significant area - one example that stands out today is the small screen - but it proved a valuable stimulus.
Don't require the rare material being mined from the mountains of South America
IBM hopes that its new tool will avoid bias in artificial intelligence
Found by calculating the strength of the material deep inside the crust of neutron stars
Can highlight in real-time the relevant regions of an image being described