The convergence of computing and telecommunications is changing lives, mostly those of the young. New developments will benefit everyone and see intuitive interfaces allowing anyone to fully reap the benefits of IT. For the immobile, travel substitution will make it easier to shop, see friends, consult a doctor or bank manager, or simply keep mentally active. Quality of life will remain higher for longer, and people will be able to take an active part in society even if they cannot get out of a chair.

A DIY society, where many people effectively opt out of the techno-centred system, placing more value on social issues than on wealth generation, could evolve into a completely machine dependent society. For the even further future, we speculate on the arrival of telepathy, with its ultimate realisation in an end to death and merging of man and machine.

Introduction
As the concentration of older people in our society increases, it is fortunate that the technology advance which enables longer lifetimes has also brought about the means to care for them. The whole field of information technology is developing quickly. The social benefits from the convergence of computing and telecommunications will be vast. People of all ages will benefit, but the special needs of the older generation will require technology which is only starting to arrive now. Through progress in simple interfaces and travel substitution, they will both be able to access the benefits of information technology, and circumvent the barriers caused by frailty. As a result, they will be able to have a higher quality of life, for longer. We explore some of the advances already taking place, with pointers to what is to come.

In this paper, we address the technologies which offer substitutes for travel, thus helping the old and the infirm. We also look at the larger potential social consequences of information technology progress. Though these changes will affect everyone, they will be particularly significant for elderly people who may find it harder to adjust.

Info-medical advances
Whilst IT can also be used to allow people to be independent for as long as possible, and informal health care can be transformed to self-help care, the question arises - is it possible to more effectively utilise the highly skilled and trained? For example:

• Can nursing practitioners take some of the load off GPs?
• Can GPs and nursing practitioners work together more effectively?
• Can consultants and specialists be more accessible?
• Could the social services be more effective?
• Could help be on tap when required?

This is all feasible today through the provision of PCs equipped with video cameras and high-resolution displays for consultants, GPs and nursing practitioners, who can share images and data, linked by mobile communications. This might seem like a futuristic, powerful personal computer (PC), but in a few years it will be available at a cost less than ?500, and be an integral part of a home entertainment and management centre!

With such technology perhaps the smart patients will be able to access medical and care help directly??Really smart patients will have access to the same information as the medical profession!

An Ageing Local Population
Average age in the developed world is increasing fast. There are relatively more old people than ever before because life expectancy is now approaching 80 years in most developed countries, and the birth rate is declining. The number of people available to support this ageing population is decreasing with each new generation (Fig 1). In some parts of the world, they expect the imbalance to decline to one supporter per two adults. In Japan, this is being addressed by developing robots to undertake the caring and rehabilitation tasks. We can expect that the rest of us will also be able to buy these robots.

Fig 1. Ratio of 15-64 year olds to over 65's

Fortunately, in much of the less developed world, the opposite trend is true, with relatively plentiful supplies of young people. Although we cannot realistically move all these people, information technology and telecommunications offers a potential solution to the geographical anomaly through the teleportation of expertise, experience and presence itself. The technologies that allow surgeons to be positioned inside the human body through an endoscope or through the use of a surrogate head peering into an incision are already with us. Before long we will see surgeons in California performing operations in London. Robots are already being used in hip replacement, brain and eye surgery. The trip to a doctor's surgery or the hospital outpatient's department could soon become an automated and remote activity.

Further medical developments include the remote monitoring of patients through electronic interfaces mounted on the body. It is already possible for diabetic and other drug and medicine dependent people to be monitored by remote computers that can administer and optimise dosage. So far experiments have been confined to hospital wards, but there is no reason why this cannot be realised via the global telecom network.

Networked Medicine
When doctor and patient meet, in the flesh or on line, medical records can now be instantly accessed. For example a recent experiment in the United States allowed magnetic resonance images to ve shared between a surgeon and the patient's own doctor many miles away. Symptoms can be recorded, and the computer can use intelligent software agents (i.e., software robots) to identify a range of causes and treatments. At the point in diagnosis when the doctor may decide to refer the patient to a specialist, the doctor can instead access the specialist's own expert system for help in further diagnosis. If the diagnosis still isn't clear the doctor can use agents to search databases world-wide for diseases etc which match the symptoms. By allowing such agents to carry diagnostic and possibly visual information, the doctor is able to get help from any medical facility connected to the network. In essence, the GP will become a cybernaut, sending agents out over the Internet to search databases world-wide to match symptoms and possible causes. When it is time to refer a patient or make a hospital appointment, an intelligent agent can check the appointment register and make a suitable appointment. If the referral is to a consultant, the patient may not need to travel. By using telepresence the GP can act as the consultant's eyes and ears and examine the patient under guidance. Specimens can be taken and sent for analysis. When the results are returned, doctor and consultant can inspect the data, images and health records without meeting and agree the most appropriate course of action.

Historically medicine has relied on such use of images in diagnosis, patient management and training. However, in the last ten years or so, the extent and importance of this dependence has grown exponentially, so much so that working with and interpreting images is central to modern medicine. Obvious examples of this are the high resolution, computer enhanced images produced by MRI and CT systems, high magnification microscopy, laparoscopy, endoscopy, and of course conventional X-ray plates.

Interpreting images fully and correctly is a specialised skill, one which ideally is widely available in every hospital and medical institution, and can be called upon immediately when required. The reality today is different. Such skills do exist but high training and equipment costs mean that they are not always where are needed at that instant. Indeed, there can be economic benefits in hospitals becoming specialised centres of excellence in particular fields. However, under such a regime patient care may involve transportation to other hospitals, or the relevant specialists making the reverse journey, all of which involves delay and raises costs, and thus works against the original economic rationale. The solution is to transport the skills to the patient in a quick, simple and low cost manner. Modern telecommunication networks already make this a reality.

With recent advances in image compression techniques, coupled with the emergence of a global all-digital telephone network, it is now possible to convey still and moving images between any locations with little if any perceived degradation in picture quality. Furthermore, such links can be readily configured to convey image, sound and computer data simultaneously. It is thus possible to deliver in real time all the information concerning a patient's condition to a remote consultant who, also in real time, can respond with appropriate guidance. In a similar manner, education would be significantly enhanced through consultants being able to broadcast interactive lectures on their particular field to many locations on the globe simultaneously.

Telemedicine, to use the generic title, is an active and growing field, embracing many disciplines in medicine and telecommunications. The majority of the evaluation programmes are concerned with conveying images that contain very fine picture detail, such as X-ray plates and magnetic resonance imaging (MRI), as the operational returns can be immediate. In order that such images be conveyed in real time, or at least within a sensible time period (e.g., no more than a few minutes), the transmission rate over the network must be correspondingly high (typically ~2Mbit/s, exceptionally ~155Mbit/s). In some specialist instances such capacities are provided by dedicated point-to-point links. More commonly, multiple telephone lines, which individually are narrowband channels, are aggregated using specialist equipment to emulate a single high speed channel. In either case, the cost of providing the telecommunication link is high. In contrast our research has focused specifically on low cost, widely deployable solutions which use only one, or at most a small number of telephone lines. This in turn dictates that suitable images are those which contain medium to low picture detail, such as conventional television and VHS pictures. In this context endoscopy is an obvious candidate. Furthermore it is a specialism which, if more widely accessible through the use of telecommunications, would experience a rapid increase in the number of skilled practitioners, to the ultimate benefit of the patients.

We have successfully conducted a series of trials of tele-endoscopy during which a number of patients at Ipswich Hospital (all of whom gave informed consent) received endoscopic examinations while the attending endoscopist was in sound and vision communication with a fellow endoscopist (Fig 2) or a group of trainee clinicians (Fig 3) at a remote location. Experiments have also been performed in which foetal scans have been viewed simultaneously in Queen Charlotte's Hospital in London and St Mary's Hospital in the Isle of Wight without loss of quality or definition.

Fig 2. Tele-endoscopy over ISDN-2. Remote consultation

Fig 3. Tele-endoscopy over ISDN-6. Remote education and training

The links used were standard BT dial up ISDN (Integrated Services Digital Network) lines. In this way consultations and expertise can be literally dialled up from almost anywhere as required. During those trials in which experienced endoscopists were at both ends, the link comprised a single ISDN-2 line (128kbit/s), whereas the group of trainee clinicians was linked via ISDN-6 (i.e., three ISDN-2 lines multiplexed to function as one 384kbit/s line). In either case the ISDN lines were terminated by commercial confravision units (BT products VC7000 and VC2300 respectively) which conveyed sound and vision at a thoroughly adequate quality for this particular application (Fig 4). The images conveyed to the remote expert included not only the conventional view from the endoscope but also the path of the instrument inside the patient's body (Fig 5).

Fig 4. Endoscope images over the ISDN

Fig 5. Path of the endoscope in 3D inside the patient's abdomen

This image is produced by a revolutionary system that does away with the need for X-rays and fluoroscans. As the endoscopist moves the instrument, the new path is immediately displayed in three dimensions. A third image relayed to the remote expert was a video camera view of the patient and attending medics. All three images were conveyed to the remote end of the link in a variety of formats that suited circumstances and requirements at the time, for example, picture in picture (Fig 6), video fade in/out, etc. At the patient end of the link the attending endoscopist and nursing staff had a head and shoulders view of the remote expert or trainees and were able to communicate in a completely hands-free style.

Fig 6. Endoscope display with an inserted view of the examination room

As well as investigating the potential of remote diagnosis, these trials have permitted a rudimentary but nevertheless meaningful assessment of the potential for local and networked education. Even at this early stage it is clear that the potential is profound. In Georgia, USA, all public hospitals are already linked up to provide these types of services, and they are planning to eventually export medical expertise to any part of the world. The scarcity of experienced practitioners, due partly to the infrequent opportunities to watch and learn directly from others, could become a thing of the past. The ability to literally dial up and observe ongoing examinations could become common place and provide an important training supplement to hands-on opportunities.

In theory, any electronic equipment used in diagnosis can be connected over a network and the display viewed in as many places as necessary. The only limit is bandwidth, and this is rapidly being overcome by fibre optics. But what about non-electronic diagnosis? Pictures and X-rays can be transferred now without loss of quality, as is evidenced in the above two trials. In fact, the only diagnostic technique not yet available remotely is touch. However, by 2015, synthetic skin with all the tactile qualities of human skin will be available. One doctor will be then able to feel what another is touching, anywhere on the planet. Technology will provide telepresence and allow doctors and carers anywhere to provide the best service possible. Part of this service may be the change from a network of GPs supported by a large general hospital, to a series of 'smart' cottage hospitals supported by advanced facilities held centrally. When GPs and fund-holders are able to diagnose and treat a wider range of problems themselves, consultants and surgeons will be able to spread their expertise over wider geographic and medical areas.

Better health care in the community and improved diagnostic skills may reduce the need for general surgery and allow the surgeons to concentrate on specialist operations or supporting doctors in cottage hospitals. The same technologies that are used in diagnosis can be used in surgery. Telepresence will allow a surgeon anywhere in the world to oversee, and eventually perform, an operation somewhere else in the world. For example, researchers at the American Department of Defense have shown that it is possible to operate remotely using crude 'remote hands' and robots. Experimental robots are currently being used in some operations in a supporting role. The AMARC programme at the University of Bristol is looking into taking robot technology into the operating theatre. Early experiments have involved robots controlled by tomographic scanners removing tumours with a precision higher than that of surgeons. One success is in Stapedotomy operations, where the surgical robot has demonstrated a performance superior to that of humans when drilling into the bone. With these successes robots are now being introduced into surgery for retina or joint replacement. However, the full benefits of this technology will not be reached until human levels of dexterity and tactile sensations are achieved, using tactile skin.

Information access
Medicine as a science is characterised by the amount of analysis based on subjective data. The diagnostic process can be long and complex, relying on techniques first used by the Greeks and needing many referrals and tests which often result in uncertain conclusions by a process of elimination. In many professions it is not unusual to spend up to 85% of your time trying to find information, 10% putting it into the right format, and only 5% making the critical decision. We need help to navigate through the growing sea of information, find what we want, access and manipulate data so we can get down to the kernel - decision and action! The necessary technologies are all under development and use Artificial Intelligence (AI) for navigation and location, Hebbian decay mechanisms for filing, and automatic text summarisation. However, there are still significant problems associated with the complexity and size of systems, data bases and connectivity expected by the year 2000. These must be overcome by the equipment manufacturers, service providers and telecommunication operators working as one.

What improvements could be made to this process?

Searching for information is a waste of human time and effort. Software "agents" are now able to roam the networks and electronic libraries across the planet, find data and retrieve it in a matter of minutes. The next step is to teach them to format data for individual applications and humans. This is only a short step away!

The single biggest impediment to realising an effective system is the huge legacy of patient data, case histories, techniques and information recorded in paper form. If all of this were available electronically, access to the global experience would be possible. Furthermore, the continual analysis and production of trend data would be feasible, also on a global scale. Hot spots, errors, corrections, new correlations and effects, new diseases, and outbreaks could be detected and be acted upon within hours rather than weeks or years. The barriers to realising such a capability are formidable, but mostly non-technical, and will probably require a further 20 years to be overcome.

What savings could be made for both the doctor and the patient if the number of tests, and delays between tests, could be reduced? How much money is tied up in paper-based systems and postal delays? How much resource is expended in merely making appointments? How many referrals result in no action required?

Smart Copier - Dumb Pacemaker
Many offices now employ smart copiers and fax machines. When the service engineer is required the machine calls for help. The same technology is finding its way into food and drink vending machines, garage forecourts, automatic bank tellers and telephone kiosks. Soon it will enter your washing machine, tumbler drier, microwave oven, Hi-Fi, TV, and computer. None of these are critical to life - but the pacemaker is, and yet at present it is dumb! As the number of artificial body parts increases there will be need to monitor both the outside and inside of humans, collect a wide variety of operational parameters, and relay them back to some 'caring computer' for analysis and action. Preventative maintenance is preferable to curative medicine!

Real Time Drugs Administration Vast amounts of time and physical space are taken up by people who currently require continual or periodic observation in hospital. Today the administration of everything from insulin to steroids requires the attention of trained medical staff. A modest amount of instrumentation and artificial intelligence would allow both monitoring and administration of the right dosage, optimised through real time bio-feedback to match an individual's body mass and metabolism. An artificial pancreas is already at a research stage, and some hospitals in the USA are experimenting with the dispensing of drugs by such mechanisms. The next step is obvious - the patients go home and are monitored and advised from the hospital machine over the telecommunications network..

The Machine as Doctor
There is some evidence that patients often prefer an impersonal, inanimate questioner as the initial stage of the diagnostic process. Talking to a computer about personal issues may be less embarrassing and more acceptable, and allow the patient to be more candid thus improving the quality of information and subsequent diagnosis. Furthermore, it has now been demonstrated that machines can already equal the human in the preliminary diagnostic phase. It is likely that they will be able to outpace and outclass us as the amount of information increases through the use of improved biometric measurement techniques. In this situation AI could be a means of reducing diagnostic uncertainty as it has the potential to consider all previous global case histories and diagnoses!

Here the primary impediment to realisation is the human - machine interface. Unless machine interfaces become considerably easier to use than today's PC interfaces, there is no hope for them in mass use. Indeed, this is another mechanism for the generation of an economically and health-wise disadvantaged class. What is required is an interface that can be driven by the lowest common denominator, whether 5 or 90 years old. Progress which is being commercially driven by Video on Demand (VoD) and other services designed for mass penetration, i.e. telecommunications, computing, games, TV, radio, Hi-Fi, and sensertainment, might provide a more usable interface.

Real Time Monitoring
Other advances will change the balance of home care from passive to active. The same technology that allows wrist-sized computing will also provide diagnostic information and location detection. People who elect to wear such wrist devices will benefit from round-the-clock monitoring. A PC in the home is able to learn the wearer's preferences and daily routine, and thus recognise when medication has been missed or the possibility of an accident and raise the necessary alarms. Combining these may allow the onset of sickness to be detected earlier and provide the means for timely treatment. Using the location detector, people who go missing can be found without the need for a police search. Alternatively, if they become ill while away from home, the device will be able to raise the alarm and inform ambulance crews where to find the person and what to ve expect.

Applications and Services
The information age in the 21st century will be dominated by choice. Users will have a bewildering range of services, information channels and entertainment outlets to select from. Our vision of this future system requires the use of agent technology to assist users to manage their personal information space. N. Negraponte (Director MIT Media Labs) has coined the term 'electronic butlers' for these types of agents. Just like human butlers the agents will have to learn the needs and whims of their owners. The distinction between home services and business applications will become increasingly blurred as home working becomes more commonplace.

The range of information channels and diversity of entertainment sources available will provide significant opportunities for service providers. Just as end users will need agent technology to help them locate, negotiate and route information, service providers will need to develop agent systems which help in targeting of audiences and users. Ideally the service provider systems should build up a profile of every user connected. Such profiling systems must have the ability to track the changing interests and habits of users and anticipate their future needs. They can then become proactive providers of new information products.

CamNet based telepresence
In many walks of life human beings operate with sensors diversely focused on a number of activities. Perhaps the most common example is that of the pilot who will talk to ground control using one ear whilst listening to his co-pilot in the other. At the same time he may be viewing activity on the runway and controlling the aircraft with his hands. A recent realisation has been that this split ear operation can be extended to the eye. It is possible for human beings to work in two visual worlds at once, the real, and the virtual .

A commercial manifestation of such a system has been named CamNet, where the operator utilises a conventional audio headset augmented with head mounted TV camera and television screen . It is therefore possible for expertise to be teleported from one part of the world to another. Specifically, this can be used for the maintenance of complex equipment, surgery, medical examinations, remote visits to new locations, loss adjusters for car repairs and many others. The principle is quite simple. By projecting the view from someone's head back to a site many thousands of miles away, it is possible to perceive the world of another human being. Instructions, schematics, fixed and moving images can then be transmitted back to the wearer, who can receive the latest information about the equipment or problem he is working on. The addition of tactile and prosthetic transmission, plus remote instrumentation, then creates the ultimate teleportation of human abilities. Expertise is then available on demand, anywhere, any time!

An obvious extension is to place two cameras at eye level on the wearer and at some distance provide a virtual reality headset. It is then possible for thousands of people to sit behind the eyes of the wearer of the cameras and enjoy a full binocular (3D) view. This surrogate head could have interesting and important applications for the instruction of surgeons, students and sports or ceremonial presentations. One person is able to gather all of the scene and relay it back to remote locations.

By the end of the millennium we will be able to 'place' the distant expert on the surgeon's shoulder by using a CamNet system or a surrogate head which allows the distant surgeon full control of the cameras, to inspect what he wants to see and point to features of interest to both surgeons. The remote surgeon only needs to be involved during the critical parts of the operation and can switch between patients as necessary. His or her expertise can be made available anywhere as required, and he or she will be able to help in far more operations.

The use of technology isn't confined to home visits by doctors. CamNet is a system that uses headset-mounted cameras to transmit images to a distant site and has uses in many industries. People who are no longer able to travel or work because of physical impairment may be able to extend their useful working life by travelling by surrogate.

Future telepresence
Ideally, communication delivers a 'being there' sensation. In the future, this will be enhanced using descendants of VR technology, but a direct link to the nervous system could make for increased integrity. Artificial senses will also be very useful, allowing people to 'sense' an environment which they could not possibly enter for real, such as the inside of a nuclear reactor. Surgeons would be able to use microbots as if they were an extension to their own body. These technologies are all under development.

Leisure telepresence could be very popular, displacing much of the environmental impact of real travel, so that people would be able to visit and experience attractive locations without causing erosion and congestion, and save money too. The same technologies which allow advanced remote medicine will help people to work, play and socialise without having to leave their homes. Social interaction could be especially enhanced when we consider the other technologies that will be around. In cyberspace, there are no language barriers, people of like mind can be found easily, and physics is custom designed!

Travel
In spite of technology advances, people will always travel sometimes. With PDAs and their descendants taking care of the user's travel plans, travelling would become simpler. We will certainly have many smart roads, and cars will probably be driven automatically on these routes to prevent congestion. Hopefully, more efficient use of resources and better energy production would see much of the environment left in good condition (even restored), making travel more enjoyable.

Virtual Reality
Virtual reality will mature early in the next century. It is likely to be an important interface tool, and may be used with lightweight spectacles, active lenses or large flat screens, and maybe eventually with a direct brain link. Resolution will be indistinguishable from reality and will be multisensory. Even if a direct brain interface is not possible, smart computers will inject extra dimensions into interactions. For leisure use, they should be able to produce fantastic alternative realities, with the smallest hint of guidance by the operator.

Computing Progress in the 25-30 year time-frame
Computers will gradually become proficient at a wider range of activities previously requiring human intelligence, including their own design, and thereby introduce positive feedback into the development loop. This, coupled with currently expected progress in nervous system connection, may accomplish a direct machine-human brain link by 2025. Enormously powerful machines, using direct brain links to join together the minds of a society, will produce a new infrastructure, which we call the Medium.

Since their general introduction in the early 1980s Personal Computers (PCs) have now completely populated business life in the western world. Many professionals now own two machines: a desk top and a lap top. This has arisen through the utility of being able to 'carry' the office to meeting places and performance/price incentives. To some extent PCs have replaced the traditional role of clerical assistants, drawing boards and filing cabinets. In future they may well simulate much of our laboratory, test and modelling space - and replace the human secretary. Today they provide a diary function, store documents, undertake word processing and interconnect their user with the global messaging systems. Impressive as these capabilities are they still demand a high degree of capability on the part of the user. This need for users of modern IT to learn the complexities of their systems in order to unlock the potential of technology is an impediment to the mature, but less of a problem for the young. Interfaces can and must be made simpler, more intuitive and open to the whole population.

Even though we can be critical of today's technological packaging and unfriendliness, it must be said that performance is truly impressive. PCs currently deliver a hardware line-up which has echoes of yesteryears' supercomputers: 100 MHz clock; 32 - 64 bit processor; 8 - 64 MByte RAM; 1 Gbyte Disk; CD, video and hi-fi sound. All this is at affordable and reducing prices. At the supercomputer end of the spectrum the computer power is truly staggering. Specialised numerical processing machines operate at 100's of GFLOPS (a GFLOP is 1000 million floating point operations per second) and general purpose machines have arrays of 100's of processors. The chances are that the performance/price ratio will continue its upward trend for at least the next 10, and probably 20, years. Such extrapolation tells us that in 2005 PCs will be delivering 10,000 MIPS (million instructions per second) and supercomputers will have performance exceeding 100 TFLOPS (i.e. 100 million million floating point operations per second).

Given this remarkable advance in hardware ability, it is interesting to contemplate the relatively static situation in software. If only it was simpler, more reliable, understandable and cost less. Why not? Software still ranks as a very immature science. We have failed to create the evolutionary environments that would create fit for purpose solutions. We are like the Romans building bridges - we have methodologies, but no true understanding.

While today's computers are easily identifiable discrete entities, in the future processing capability and intelligence will be ubiquitous. The vast majority of processing devices will be transparently integrated and unidentifiable as individual machines. However, there will still be a place for machines dedicated to providing computing and communications capability, but these will be very different from, and far more powerful than those of today.

Fig 7. Computer -v- human evolution in abilities

Computers have long since surpassed human digital processing capabilities, but we do not yet consider them intelligent (whatever that means!). Indeed, today's supercomputers are thought to have the intelligence of a chicken. However, many of the functions that we associate with human intelligence will be capable of machine implementation by 2010 (Fig 7). Examples are the use of natural language, equal or better senses including vision and recognition, understanding text and graphics, and machine use of human-like memorising, creativity, recognising and learning. Already, computers can beat most grand masters at chess. Some computer architectures and mechanisms will be based on the human brain with a full range of high fidelity human-like senses, and pre-processing in biosensors where appropriate. With neural networks (or other technologies) in principle it is not necessary to know the algorithm used by a processor in order to instruct it to solve a problem or make a decision. Our 'cleverer than human' machine would be able to 'understand' a problem with the same level of guidance as neded for a human, and would 'think for itself' as far as solving it. In many ways they will be smarter than humans with their interface to the real and computer generated worlds centred on these new and enhanced senses. By about 2025 it may even be possible to link directly to the human brain, certainly links to the peripheral nervous system will long since have been achieved.

Technological Incest
We can expect these machines to have a significant impact. For example, since it is conceivable that they could design computers better than those which humans would be capable of, the rate of development of computing would accelerate, each generation designing ever better computers at an ever faster rate, perhaps eventually inventing concepts that would be incomprehensible to humans. A typical example of recent human displacement is in Printed Circuit Board design. A human operator may take days to produce and verify a design that a desktop computer can accomplish in hours. As speeds and artificial intelligence techniques improve, machines will take over more and more human tasks such as design.

Computer Performance
At present digital computing speed is increasing by a factor of 100 per decade (Figs 8, 9). There are no reasons to expect any decline in this rate in the foreseeable future, but some reasons to expect it to increase. Indeed, we could argue that positive technology feedback could give an additional improvement factor of 10 per decade over that resulting from human research alone. Thus, the overall improvement rate could be 1000 times per decade. The 2030 super computer could be 1 million times more powerful than the brain. However, there remain many areas unique to humans that these figure do not (and cannot - yet!) take account of. By the same line of reasoning the PCs of 2030 may be equivalent to 10,000 human brains. They will not be recognisable as PCs by then of course.

Fig 8. Projection of supercomputer speed

Fig 9. Projection of memory 'chip' capacity

Much of the work carried out by computers of 2030 may involve analogue processing and computation. Since the world is mostly analogue, it seems reasonable to expect that analogue techniques will be used to interface with it and to deal with the problems arising from it. Indeed, many problems may only be soluble using analogue techniques. Computers will use a mixture of processing techniques.

Personal devices
Everyone would have a personal device with enormous processing and communications power. It would be capable of anything we currently dream of having in a computer. The user could communicate with it by natural voice, body language, even by thought if a direct brain link is accomplished. It would play an essential part in the user's life, acting as his agent for just about everything. It would have full access to the Medium. We will call this personal computer a PAL (Personal Agent for Life). It would connect to the Medium to offer full functionality, but be capable of many functions without direct connection. It could read 100 terabytes smart cards and the few remaining dumb storage media, so still offer all the functionality of the electronic book.

Whether the PAL would be the user's only device is uncertain. For example, it might not yet be possible to include all the required functionality into a small, portable device. If not, then the user would still need a larger base at the office or home into which he would plug his PAL to achieve full Medium functionality. The PAL would thus be able to link to the Medium when necessary. Outside, this would rely on mobile communications, making good use of the fixed mobile network too. It would provide access to any communications or processing the user might require.

PALs would store personal information and preferences, and would be able to act as an agent wherever people need to interface to the outside world, taking care of financial transactions, scheduling, information collection, entertainment organisation etc. People would be liberated to relax and enjoy the finer things in life. Those wanting to escape from it all could go for a walk, leaving their PAL behind to act on their behalf when required. More likely, it would be taken along to act as a guide, interrupting only if permitted or requested. Being capable of conversation, it would make an ideal companion - talking only when appropriate. It would be indispensable.

Impact of powerful computers
The impact of such powerful computers will be staggering. They could make better decisions than humans, with instant access to all current knowledge via the world's databases, and enough storage and processing capability to use it. Human roles would change dramatically. The following are a few of the possible outcomes:

? Companies would have a more chaotic and amorphous structures, many existing only in the infosphere, with no physical or centralised base. They will be transformed, reorganised, merged or wiped out to the extent that none of today's companies will still exist in recognisable form. It is possible that the whole economy would tend to stabilise under automatic control and co-operation at the highest levels.

? Knowledge would multiply rapidly, advancing every branch of science, medicine, and economics, assuming we can persuade computers that these are of interest and worth tackling.

? Computers could be made very friendly. On the other hand, the shear power of these computers will leave mankind intellectually dwarfed, and there will be vast pools of knowledge that only computers can understand and use!

Networking Progress
The same technical and commercial forces that shape the computer scene also affect telecommunications. Digital transmission and switching now provide near perfect voice communication across most of the globe. Mobile systems allow people to enjoy unprecedented freedom as to where they access networks. Satellite technology enables telephony even whilst in the air or at sea. A number of voice activated services are available already and this will increase rapidly.

Other technology leaps
With technology feedback, we can expect that research and development in almost every field would have made huge strides by 2025. In areas limited by theoretical extrapolation and lateral thinking (bringing together superficially unconnected ideas to yield a useful result) such as quantum mechanics and mathematics, we may see the largest gains as computer intelligence really makes a mark. Other areas may be limited by the time and resources (or ethics) required to build experiments to find missing data. Indeed, physical resource limits may be a significant problem. In the 2025 time frame:

? New materials will have been discovered and new engineering techniques developed. Currently novel technologies such as nanotechnology will mature quickly. Advanced materials and technology may allow lightweight cybernetic additions to the body to enhance strength or provide protection. Artificial exoskeletons may become more versatile.

? Body 'spare part' technology, together with the new range of sensors, could realise the 'bionic man'. Such body 'add-ons', perhaps linked to standard interfaces, may become everyday accessories, allowing people to select and optimise their abilities for the day when they get dressed.

? Safe genetic engineering would have a great impact. Genetic models may allow genes to be designed and simulated before building custom organisms. Gene therapy would help to avoid or alleviate many ailments. Many current diseases would have been eradicated. Life expectancy will be higher due to better medicine and social care.

? Ethics permitting, parents could specify the properties of their children. Genes could be selected from a gene bank and a custom zygote produced.

? Even if a full brain link is not available by 2025, direct nerve links certainly will be and devices would be available that provide new 'senses'. We may see tactile devices which sense surface properties much better than our fingers, ears sensitive to a wider range of sound, safety devices which can detect radiation, navigation aides which allow people to 'see' magnetic and electric fields and many others.

Harnessing the machines
Interfaces

It is a sad fact that a large proportion of the adult population have difficulty programming a video recorder. Many people are still afraid of computers. This is not due to poor intelligence and ignorance, but poor interfaces. The real world provides our senses with a very rich experience and we are well equipped to deal with it. By comparison, the best current technological alternatives are pitifully impoverished. Human kind was not designed to cope with spread sheets, the written word, keyboards and small screens that only present a partial picture of a wider activity. Developments in artificial intelligence, visualisation, virtual reality and telepresence will realise new capabilities. Imagine a virtual reality interface with your visual cortex flooded by information from spectacles or active contact lenses, each augmented by directional audio input, tactile gloves and prosthetic arms and fingers that give you the sensation of touch, resistance and weight. Imagine also the prospect of a surrogate head that is either machine or human, and which can allow you to be teleported into environments anywhere on the planet with great accuracy and reality. This leads to the euphemism that is already being realised in actual applications:

?what you see I see, what you hear I hear, what you feel I feel?!

Alternatively, contemplate the convenience of large visual displays with high definition in two or three dimensions. People could appear in full proportion and true colour, with a voice that emanates from the lips in a distortion free and convincing manner. All of this would add further to a feeling of being there!

Current access to computers via mice and keyboards is archaic, designed to slow down the user; it requires a substantial investment of effort to achieve even moderate competence. Graphic tablets, touch screens, pens, joysticks, data-gloves and pointers provide partial solutions but do not come near an intuitive interface, which must be speech based. Speech recognition is advancing at a pace whereby we will see, in the next few years, the much more natural interface of being able to talk to the computer and have it respond to simple commands. By the beginning of the next millennium, topic specific applications with large vocabularies will be in use. Computers with understanding, able to learn as we do, will come when major advances are achieved in language modelling.

This is also true of information. Early experiments with graphical data and emotional icons in this context have revealed a number of interesting advantages when linked to Artificial Intelligence (AI). Reaching out for information that reacts in a humanised way as being friendly or aggressive, defensive or nervous, allows participants to be steered through a decision making process with realistic cues abstracted from real life. This is in complete contrast to most interfaces that have been engineered for the convenience of the technology rather than the human user.

While all knowledge transfer between humans is accomplished via speech (whether vocal or written), it is images and visualisation that provide the key to understanding and (probably) many of the breakthroughs in science and engineering. Our view into the world of computer generated imagery has advanced slowly. Based on the cathode ray tube, or liquid crystal display, we look through a tiny grainy window compared with that which our eyes normally provide.

It has been known since the 15th century that humans were able to project themselves into an ?abstract' world, with the Linear Perspective Panels of the architect Filippo Brunelleschi. Looking through a peephole in the back of a painting at a reflection in a mirror in front, the person viewing the image believed that they were seeing the actual scene. The majority of humans are able to move from the world of abstract images created on the computer to the real world, with consummate ease. The concept of the virtual reality (VR) headset, first envisaged by Scot Fisher for the NASA moon programme, was to enable astronauts to experience walking on the moon. Stimulating the auditory and visual senses, interaction with the computer world is achieved by head tracking and data gloves. Since that time, the major changes have been weight reduction and in the cost. Driven by the huge market for interactive games, Nintendo will release a $200 system onto the market by Christmas 1995. The modern headset is still cumbersome and awkward to adjust, and using CRTs for the display, it needs to be connected to the computer via an umbilical. The next step, already in development in the USA, is to provide the video screen mounted on the lens in a pair of spectacles. In the future we may see the technology transferred to the contact lens, and means of interacting with the eye, so that when we change the position of the eye (e.g. looking up or down), without moving the head, the object of interest comes into focus.

At the other end of the spectrum, wide angle (>160?) projection of images offers an alternative realistic interface, with real time textural rendering of surfaces carried out by top-of-the-range workstations. Currently in use for high value applications such as flight simulation, today it is possible to obtain a pilot's licence for a Boeing 737, without ever taking off. The cost of these systems is prohibitive (of the order of $200k for the screen and display), but technologies for projection will provide better resolution and bring prices down dramatically. For example at Texas Instruments a micromechanical chip has been developed which contains millions of microscopic mirrors. Under electronic control the mirrors can tilt and direct an incident light beam onto, or away from a screen. Such a device could easily fill the largest (1152 columns and 2048 rows) HDTV screen.

It is in such abstract worlds that not only will understanding through visualisation be enhanced, but, by stimulating the power of the human brain, new theories and concepts will emerge to extend man's knowledge.

Full Interaction
The purpose of telecommunication is to effectively teleport us somewhere else. Unfortunately, technology difficulties do not yet allow this to be very effective. Today we use the telephone to teleport over vast distances, but the sound is low quality and directionless, our voices are muffled, we are blindfolded, our noses pegged, and straight-jackets deprive us of any tactile information. Not surprisingly, people generally prefer to meet in person where possible. The same is true for television, where the screen and camera effectively make a small hole in the blindfold. Extending our sense of touch is also feasible, as the time for information to travel from our fingertips to our brain (~30 ms) is of the same order as the time taken for a single photon to travel from the UK to the USA (~ 30 ms) on an optical fibre. Given that this time delay is dominated by synaptic processing (~30 ms), and our innate ability to accommodate delays of 100ms or more (e.g. riding a bicycle), then an extended hand across the Atlantic is feasible. The ?feely gloves? and ?prosthetic arms? currently being developed will allow the VR participant to reach out, and not only grab, but feel and then react to the tensions and forces, the tactile qualities, size and viscosity of entities. Elaborate data suits allow sensory information to be communicated for the whole body. Even smells can be sampled and recreated.

Current research suggests that direct sensory stimulation will be practical in the next decade or two, which should eventually improve the feeling of teleportation. These developments will also help full-sense prostheses. So while current technology means that telecommunication offers a poor substitute for being there, technology is advancing quickly in all the relevant areas to make it more effective. While people still travel to see each other and other places, telecommunication will become an increasingly acceptable alternative as the quality and richness of the experience improves.

Ultimately we might envisage that millions of people could be teleported to a single seat in Yankee stadium or the Wembley arena. A single optical unit (rotating camera) placed in one seat could gather all the optical information and teleport it to thousands of locations. At these independent locations the wearer of a VR headset would be able to move and look to front, back and side and see all of the scene around that seated equipment. In the future we might be able to experience the full panoply of the Olympic Games, without leaving our living rooms.

Decision support (AI)
Currently, medical research is producing an exponential growth in information. New drugs that doctors need to understand are being released continuously. To benefit from this information explosion, doctors must be able to find the facts that are relevant to the task at hand. Since the early days of computing, expert systems for medical diagnosis have been a main target for Artificial Intelligence (AI) researchers. Whilst the early academic systems proved as accurate as human doctors they suffered from lack of medical confidence and accountability. However, more recently the growth in information and improvement in AI techniques, such as knowledge elicitation and representation, has seen systems surpass their human creators in accuracy. Whilst it may be a some time before automatic diagnosis is the preferred option, such systems will at least give decision support and on line reassurance in diagnosis and treatment for doctors. The artificial expertise to diagnose and treat a wider range of problems without the need to refer patients to consultants is only a short step away.

Agents
During the last decade research has demonstrated clearly that expert systems can automate many network management functions. Both rule and model based expert systems can diagnose faults to a very high degree of precision. In addition, expert systems are making inroads into areas such as network planning and system design. Research into distributed network management clearly shows that distributed automation is feasible. In fact the kernel of the AI systems which power distributed management have been given a name - agents. This term is often misused. In our view the term agentshould be reserved for a particular kind of AI system - where negotiation is a central part of the software functionality and where the agent can say no to a request.

An agent is a self contained software programme which has two parts:- A generic part which performs Communications and Co-operation with other agents and users, including negotiation, and a role specific part which carries out specialised functions associated with its task. Because agents are complete software systems that conduct a specialised task they are ideal candidates for distributed systems. Furthermore careful design of agent software could well allow significant reuse - in effect agent software could become an off the shelf product which will help to contain the software mountain.

Merging with the machines
There is no doubt that our brain of some 10 11 neurons is a remarkable machine. Unfortunately, it is one that will see no significant enhancement in the next several thousand years if left to natural development. In dramatic contrast the electronic computer is evolving at a rate that doubles its ability each year or so. This will see desk top computers 10 3 times more powerful in ~10 years, 10 6 times in ~20 years, and possibly 10 9 times in ~30 years. A sobering thought! Interestingly, the telecommunications, electronics and IT sectors are the only ones able to sustain such exponential growth as all other human activities most definitely face physical limits that are all too visible. IT and communication can continue the annual doubling process for at least another 30 years on the basis of the technology we already know about and understand.

Fig 10. Potential replacement of body parts as a percentage of body weight

Much of the human body can already be replaced or substituted by artificial parts (Fig 10). Parallel advances in medicine materials science, electronics and biotechnology are bringing closer the day when we can have a practical connection between our bodies and our technology. We can then envisage the ultimate extension of mankind's sensory systems along optical fibres to distant places to participate in new experiences that will change our perception of the world and the way we work and live. We will then be able to have new experiences, both real and artificially generated. To do this over any distance fundamentally requires a broad bandwidth and minimal delay.

An end to death?
Advances in biotechnology continually increase life expectancy, but at some point, we all have to die - or do we? Suppose that when our loved ones are dying, we could transfer consciousness from their brains into a super computer. We may object to the thought of transferring a loved one into machine form. However, many of us would do anything to maintain contact with our loved ones, and may consider that physical manifestation matters less than the magnitude of potential loss. Even if such technology arrives too late for our generation, it will certainly be around within the next few generations. At some point in the development of our technology it is almost certainly going to be the case that we will be able to transfer human minds, and beings, into some machine form. At our present rate of progress less than 20 years will see super computers with an equivalent processing and storage capability to the human brain. Within 30 years the development of such machines should have reached our desk top or even pocket. Whether such machines will be capable of supporting new life forms, or imbibing existing ones, remains to be seem. Probably the key challenge is understanding and accessing the human mind itself. We do not understand how the human brain works as we have not yet unravelled its unbelievable complexity. Perversely, with the assistance of super computers, it might be possible for us to create sufficiently good models to achieve a full understanding.

What kind of world will it be where no one ever has to die, or for that matter suffer unbearable sickness and failure of health? A world where we can live inside a static machine, communications network or inhabit some android or robotic form? The answer of course is totally different from what we have experienced so far. While we might expect severe culture shock, moving into a machine world, perhaps the coming technologies will prepare us well. Probably the only inkling we are able to gain of what might be possible, is currently available through the use of virtual reality. Experiments with the incredibly shrinking scientist, able to wander around atomic structures and feel the forces combining the nucleus and electron by merely putting out their hand at one extreme and at the other, our ability to wander throughout the galaxies and tamper with the influence of black holes in a virtual cosmos. Being able to go anywhere and do anything at any time and any place will soon be with us through the magic of telepresence, VR and tremendously increased computing powers and an ability to communicate over vast distances at virtually zero cost. The next logical step is the transfer of mind itself into machines away from all the limitations imposed on us by our present biological form and that of our currently limited electronics. This may be just another small step along the telepresence route. Perhaps our only chance of making full use of such technology is to become an integral part of that technology itself! The advantages are almost unthinkable.

Direct brain interface
Direct stimulation of nerves and the interpretation of signals from nerves will be possible by 2010. Artificial sensors will exist then, opening the possibility of enhanced 'add-on' senses for humans. A full interface between the human brain and a computer may be possible by 2025. It is already possible to control simple actions on a computer simply by thinking, though not yet for a computer to stimulate thoughts in the operator. We have already addressed the effective immortality possible as a result of such a link. Some other mind-boggling results of this technology are :

? Human brain enhancement by linking thoughts in the brain with those in the computer, and using the computer's superior capabilities. Linking to a powerful computer may enhance intelligence, consciousness and awareness. Whole new areas of experience would be accessible to the human.

? Telepathy becomes possible, by upload, transmission and download of thoughts (Fig 11). It may become possible for two people to share the same consciousness simultaneously, opening up whole new possibilities for personal interaction

? By creating physical experiences and sensations, the computer would be able to re-create any life situation for the user. This alternative reality would give the sensation of actually being in an environment with much greater 'reality' than can be offered by mere holograms or conventional virtual reality. Therapy would be enhanced, as people could be put through 'real' experiences without danger, in order to gain confidence. Punishment for criminals could be revolutionised by the same token, e.g. the burglar could experience the trauma of being burgled!

? Instant training and education. Learning and education in the traditional sense could be obsolete since knowledge and skills could be available via the computer. This may tend to 'equalise' people.

? What would the result be if all of these computers are networked while people are linked to them? A global consciousness!

Fig 11. Communication via the 'medium'

Clearly the above effects would have profound implications for both individuals and humanity. Personalities in the society could be pooled, with artificial personalities being created to add to the options available. Users would be free to choose personality, memories and skills from the common pool, so there need no longer be a permanent mapping of mind onto particular brains. The concept of the individual as we know it may cease. For some people the first time they upload their personality onto the machine and download a another personality may be the last time their own personality is used. They will effectively have died as an individual!

Future social scenarios
Virtual Communities

In addition to assisting in people's physical needs, IT might be able, in some part, to satisfy and assist in their emotional needs. Networks allow virtual communities to be created anywhere on the planet without the need for introductions (Fig 12).

Fig 12. The virtual community

Special interest groups already abound and discuss anything of common interest and exchange experiences. People ranging from scientists to prisoners and the homeless all access and use the Internet. There are already many bulletin boards for a large number of interests. When people have access from home and are familiar with the concepts, then the use of videophones with life-sized and life-like images will create larger and better informed groups. This in turn will shrink time, distance and ignorance still further. Perhaps there will be a new sense of belonging, new social groupings, new nations in cyberspace!

Virtual businesses and remote working
We have addressed medical uses of CamNet technology. Those concepts can be extended to other disciplines including the repair and maintenance of oil rigs, electronic and power installations and even activities in the home. Being able to teleport experts to your location, and have them guide you through the necessary steps to affect a solution is only a short step away. A key advantage is that it allows people's expertise to be used without the person actually having to travel, permitting acceptable access to expertise from the infirm and the home worker. As costs fall, virtual businesses could be formed providing expertise on demand across wide fields. For those for whom travelling is difficult, this could extend potential working life.

There is a general trend towards shorter contract working and more dynamic employment. In addition, older people experience discrimination and difficulties in finding work. Fortunately increased use of IT and the trend towards virtual businesses may help the situation. Firstly, in some industries, decision makers are having to invoke IT to become 10 to 100 times more effective. Delay is being removed by information networking, and understanding is being increased through visualisation using simulation, emulation and animation. Many high tech companies have now gone paperless throughout their management structures. If you don't have and use IT skills - and that does NOT mean a secretary doing it for you - then you don't have a job!

Secondly, just as information technology allows groups of people to socialise, regardless of geography - it can also allow them to work together. As well as geography, the age barrier can also be conquered. People can be employed according to any criteria and any company wanting to avoid older people will still be able to do so. However, top driven companies will not be the only ones. The same tools and markets are available to bottom-driven virtual companies. Today, taxi drivers form groups with a single contact point, but are otherwise self-employed. Many virtual businesses will spring up in other areas, with agents bringing together groups of people who can work towards a profitable goal. Some of these will simply allocate subcontracted work to the best person for the job. Others will be more sophisticated alliances, exploiting market opportunities and linking many people with very diverse skills, on a highly dynamic basis. In this market, decisions will be based more on suitability for purpose. Rewards will be shared among participants - and the agent. On the network, no-one knows you're 75!

This will change the nature of business dramatically. Companies best suited to the market will thrive, others will vanish - no room for dogma, only results count. People will compete on an equal footing. Those who have well equipped home offices with marketable skills, and willingness to exploit the benefits offered by IT, will be much sought after.

The Experience Society of 2000 - 2010
The vision of mankind's development from hunter gatherer, to farmer, to industrialist and more recently to information worker is now well established (Fig 13). It is now worth posing the most fundamental of questions - what happens next? It is already reasonably easy and clear to see where we are going - towards the experience society. After all, access to information is just the first step towards accessing direct experience via the network. Television and the telephone effectively extend our optic nerve and our acoustic senses over vast distances so that we can see and talk with other people. Research on tactile gloves, prosthetic limbs, tactile and prosthetic suits are already reasonably well advanced. Through virtual reality we might also soon have human reactions to sight, sound, touch and position extended to remote locations in the real world and/or computer generated worlds. Humans are fundamentally able to absorb about 1 GByte of data per second. With a modest amount of compression this can be reduced to 1 Gbit/s. For the moment we should treat this as a reasonably safe minimum bound on the bandwidth we should aim at delivering over the information superhighway.

Fig 13. The ages of mankind

What about delay? If you stick a pin in your finger, the time period between the point puncturing your skin and the Ouch! emanating from your mouth is around 30 milliseconds. This happens to be about the same time it takes for a photon to travel from London to New York over an optical fibre - very convenient! This means that any prosthetic capability will incur a doubling of the delay along our limbs through our visual and auditory cortex. Interestingly we specify hi-fi systems to have a low end cut off frequency of around 30 Hz, and flicker is imperceptible on TV, films and video for picture refresh rates above 30 - 50 Hz, depending upon content. The time period associated with these frequencies is again approximately 30 milliseconds !

For the experience of a top surgeon, scientist, teacher or artist, or being at a sporting event, library, museum or such like to be shared by a large number of people, optical fibres and network transparency are the only technologies that can achieve this. The TV, radio, computer, telephone and systems of today are fundamentally limited - at best they can only provide a narrow window, and are incapable of delivering direct experience at all!

The DIY society of 2010 - 2015
The DIY society is a concept of two parallel societies, one wealth generating and highly technology centred - the other an 'opt out', DIY society, where other values rule. In this world, a handful of global organisations dominate industry, highly efficient and dynamic, with no allegiance to any particular nation. Their business decisions have a profound affect upon ordinary people's lives, to whom they have little, if any accountability.

Only those willing to be responsive, to adapt and offer 'skills on demand' find work, which is mostly on a part time or short term contractual basis. Those out of work or retired constitute the developing DIY society in which they participate in activities that indirectly support themselves, their dependants and the community.

Many people are unemployed with no wealth generation ability and little money to spend on leisure, while those who have work are working as hard and as long hours as ever. Work is uncertain and threatening, but an exciting environment for those who can handle it. Workers who are adaptive and prepared to swap and change jobs and pick up skills 'on demand' can maintain their position in the job market. Continuous skill acquisition is needed with a broader, shallower skill set than the traditional single skill work profiles.

Those who are retired or made redundant from mainstream employment undertake activities that maintain the quality of life that they had attained. i.e. repairing the fabric of their community, self education, self proficiency etc. Occasional work provides a top up to venefits. No longer can they depend upon regular work in the traditional sense as the basis for their livelihood. The main reward for many of these activities comes in non-monetary indirect forms, e.g. better living environment, self fulfilment and quality of life. People are now judged more on the intrinsic value rather than the monetary value of their activities.

The Global Web offers an alternative trading place for skills and products. It offers small businesses world-wide reach and operation in free and open global markets where they are able to access resources and markets irrespective of geographical locality. Many of these businesses form multiple strategic alliances on the web to meet new market needs. The concept of 'Virtual Businesses' has taken off, with the same business unit chopping and changing activities and alliances to benefit from 'windows of opportunity' presented by a dynamic and often unregulated market place. They are difficult to trace and difficult to tax.

The needs of communities are filled by small co-operatives. Many use local trading systems as well as traditional ?national' money (causing taxation difficulties for governments). Work activities which had been priced out of existence are once again viable.

There is a marked increase in global wealth but a marked decrease in its distribution. Society as a whole is more polarised than ever, with those in work very much wealthier than those unemployed. Large chunks of society have formed islands of virtual self sufficiency, exchanging skills and care with a bare minimum of co-operation with the rest of society, mainly those in work. These are the DIY society. Responsibility for support has been redirected from government to families and members of the local community.

There has been a backlash by some people against the global organisations perceived to be the cause of social and environmental problems. Increasingly people refused to co-operate with the large retail distribution companies that had come to dominate sectors such as food distribution, and opted out. They also found that within the community they had the skills and resources to develop alternative production and distribution activities.

For some this meant a return to traditional values and austerity. For others DIY communities developed which harnessed the best of the technology to provided a secure and self sufficient community. This greater self-sufficiency within families and the community became the major driver for the further decline of traditional industries and the undermining of the traditional economy.

In countries like the UK the proliferation of networks and competition between them meant that communication costs became very low and profitability marginal. Many networks became unviable and there followed a period of consolidation and rationalisation of networks. Because of its importance to local communities much of the local infrastructure passed out of the hands of the big global companies and into the control of local communities, at a knock down cost.

These secondary networks have become fundamental to the survival of people and communities. Education and health care are economically provided over the networks. General knowledge that supports the DIY society is accessible whilst difficult problems are solved by remote expertise or shared experience. To help bond the new communities there is greater interest in local cultural productions and participation in local events across the web. But this participation can be global as well.

The machine dependent society of 2015 - 2020
It is possible that the following scenario could evolve from the DIY society as machines play an ever increasing role, and many of the stresses of that phase are resolved.

Powerful computers have gradually become better at inference. As a result, they have become more efficient at a greater variety of tasks, including those associated with actions based on deduction. Previous generations have had increasing involvement in the design of the next, and now all machines are mostly designed by their immediate ancestors. The business world has been so deeply affected by smart machines that the economy has been re-ordered, with a 'black box' infrastructure and the nature of human work and rewards fundamentally changed. The various 'systems' are now 'fly by wire' and so complex that only machines can fly them. People are having great difficulty in adjusting to the new technology, but business interest and global competition prevent any pulling back on the reins. Without the machines, a society simply can't compete. There is no going back.

People no longer draw their identities from their work, leaving many people confused. For many individuals, this is a big problem. Their self esteem has suffered since they were strongly valued in the old system. More specialist machines have to 'talk down' to people to communicate information. Thankfully, the human machine interface is more 'cuddly' than it was in the 1990s, to the extent that some people now feel frustrated by less accommodating humans, resulting in interpersonal friction.

A new economy now exists. The old industrial, agricultural, information and services sectors are now part of an automated 'black box, producing higher output (and thus wealth) with lower input and little or no human involvement. There are great benefits in terms of extra wealth and greater human support from a machine dependent society, but the industrial and economic fabric is so dynamic that it is unstable, requiring constant intervention by machine to prevent collapse. Humans could not react fast enough to manage it. Inference allows systems to be designed, built and managed which would be beyond human control. If the machines all failed, humans would be unable to run these systems.

However, not everyone is happy. There are great social stresses. 21st century Luddites resented control of their world transferring to machines, regardless of benefits. In many areas there has been significant unrest, but most people don't want to go back - the benefits are worth having.

Conclusions
We have discussed the impact of developing and expected computer and telecommunications technologies on society, especially its older members. While our society will become much older, the technologies to help the older members are already being developed. Travel substitution will allow us to work, play and socialise from anywhere. Telemedicine will revolutionise the way doctors work and communicate with each other and their patients. The same technology will allow experts in all fields to share their expertise worldwide without having to travel, maximising their contributions.

As computers become smarter, they will become indispensable tools to help us deal with the ever increasing bombardment of information. They will also become faithful and competent servants in the home, managing our everyday affairs on our behalf. They will gradually replace people in more and more jobs, taking higher roles until society becomes totally dependent on them. Business will be radically changed, with global virtual companies, managed mainly by machines, and people hired on short contracts. Those who can adapt to very quickly changing demands will do well, but many will not. At some point, many in society may feel we have gone too far and opt out, forming a parallel 'DIY' society, with people working more for each other and for the benefit of the community, with less focus on wealth generation. We will have a machine dependent society.

Technology feedback will accelerate the improvement in machine intelligence, expediting the machine dependent society, but also helping to solve major problems and advance other technologies. We will soon make progress connecting machines to the human nervous system. Initially this will give artificial senses and better prostheses. We will also soon see thought recognition being used in the computer interface, evolving over a couple of decades with technology feedback to a full direct brain link. Then it is possible that we will be able to communicate telepathically with machines and each other via the network. A little more progress, and we will be able to download our minds into computers, making death a mere state change. We need never fully lose a loved one again. By then, we will be well used to using the computer as a mind extension, with access to all human knowledge and sharing consciousness with each other, so losing a part of that by having our bodies die may not even be such a shock.

All this is within the scope of progress in the next few decades. The future is not so far away, and it will be very different from today.