John Lindsay, Reader in Information Systems Design, Kingston University, UK
Transcript of a lecture given to the BT/UKC/CIS Strategy & Planning Unit Conference at the Ramada Renaissance Hotel in Brighton, on January 16th 1989.
If it's any consolation it's 27 minutes to 10. At this time Placido Domingo is entering the hall on the shields of about 30 Nubian slaves and Aida has got another 1 1/2 hours to go. And you've paid 50 or 60 quid a head in order to be there. I'll try and be a little more brief than Domingo because the penalty for him, you will remember in Aida, is he gets locked up in a tomb and the death alone takes the last half an hour.
There are three really fundamental principles of speaking after dinner. The first is to be humorous, I've done that; the second is to be brief, I am not going to do that, and the third is not to say anything which might offend your audience. So I thought it would be suitable to start off by introducing you to a tabloid journal that you might not all come across, called Capital Gay. Because in the last year, something very interesting has happened on pages 18 and 19. These are adverts for telephone dating services. One of them according to a very small study has generated 1 1/2 million telephone calls since it started. Their capital costs consist of putting the advert in Capital Gay. There is a difference between peak time charges of 38p a minute and off peak charges of 25p a minute. The only difference, I can imagine, is that the 12p in those differences has gone to British Telecom. Those of you who are nasty and competitive will be pleased to know that not a single one of them uses 'you know who's' telephone lines. The only other thing they have to do, is employ somebody called Arthur, who doesn't appear to be a computer, but if you use a few words which he can recognise then he beeps in and disconnects you because either 'Nanny Telecom' or somebody outside British Telecom in 'Nanny Government' decided the last thing that was to happen in Britain, was that people were to use naughty words over the telephones because after all it's illegal to use naughty words through the mail.
Those of you who have an international perspective on telecommunications will know that if you do a classic MBA study of Prestel and Minitel, Minitel has been fantastically successful. You might also know that the bulk of Minitel's trade appears to be in the 'ladies and men of unseemly and improper virtue' who use on-line code at 300, baud and can you imagine trying to type it all in, in order to form salubrious relationships late at night? There must be lessons in here somewhere or other about the nature of information and the potential market in which British Telecom and its staff have a capacity to generate a successful niche and add what I'm told is called value, to their commodity.
At that point we will end MBA-speak and go into what we consider to be really the fundamentals of information. Because it seems to me that if we are looking at the sort of market that information is moving into in the foreseeable future, I make four quite distinct market niches. I'm sure they're only a historic accident, they won't last, but I think they're useful in terms of having some history and perspective on what we are on about. And I divide them up as being information, networks, processes and services.
The first one, information, is really historically the province of people like librarians and publishers, from the days of Asurbaniphal organising the Lands, Registration and Taxation system of Assyria on tablets of clay, little clay marks, organised in a particular order, capable of being retrieved and so on and so forth. There has been a tradition in the organisation of information and the function of an information market. Predominantly historically, the function of librarians and publishers probably, with the possible exception of railway timetables and telephone directories.
But there's an idea there of an information which we have some sort of idea about. The basic idea of information is that you have messages; that messages contain data, that that data acts as evidence on the basis of which people test hypotheses, on the basis of those hypotheses they take decisions, the consequence of which is actions, the consequence of those actions can be measured, on the basis of that measurement the information can be valued.
So information evidence, the basis of which is the testing of hypotheses, the consequence of which is the taking of decisions, the consequence of taking those decisions is action, which has a measurable impact upon the world which either results in the hypotheses being validated or being devalued.
The market within which that exercise takes place, we have traditionally recognised as saying that every message has an author and it has a reader. The reader and the consequence of taking those decisions and those actions, in turn becomes an author who initiates a new series of actions which in turn has a new reader. We recognise that if we want to bring that slightly more down to ground that basically every message has a producer, that every message has a consumer, and that the process of getting the message from the producer to the consumer is the process of communication which takes place in a concrete medium.
For the first several thousand years, it was the voice to the ear through the air; for a couple of thousand years it was clay tablets and bits of rolled up grass and the odd skin of the poor sheep or enemy who happened to be lying around the place; for a few hundred years, it's involved printing presses. And for the last 20 or 30 years it's involved really the third substantial revolution in information which I'll come on to later on.
At the more abstract level behind this producer and consumer engaging in communication, we have an intention, the author has an intention in producing this message and the reader, the consumer, has an intention of consuming the message, not necessarily clearly stated, it's not necessarily conscious, it might not have been written down on tablets of clay beforehand, but there is some sort of intention there. And this communication takes place within a context. It is the context of the production and consumption of the message which is creating the context of meaning of the message.
In order for this meaning to occur, there has to be a grammar and there has to be a vocabulary. There has to be a set of rules by which communication can occur and there has to be a lexicon, there has to be a set of data elements which are capable of being translated. And these in turn have to function within a market because every activity of production and consumption involves human labour and because they function in a market there has to be some mechanism of pricing. The consequence of them functioning in a market and having some concept of pricing, is that inevitably the production and consumption involves some process of intellectual property. The author and the consumer consider themselves to have some rights over their products and the activities of their consumptions and under some circumstances, these activities occur in privacy, in secrecy, or there is some intellectual property, there is some different pricing mechanism according to the mechanism of being able to make access to this information conceivable or to be able to protect the information from being accessed.
Now those are some rather basic primitives that we can say are generic to this concept called information which makes it behave in the market rather differently from most other commodities that we have to concern ourselves with. So we are going to, out of those primitives, produce some ideas about the way in which the information economy might actually be seen to function.
Now then, lets move on to this idea of networks, because they really are, in a sense, a little bit more complex. Because as I said a couple of minutes ago,it's this convergence of computing and telecommunications which really is the interesting thing we have to be bothered with. It's changed the relationship of space and time as far as information is concerned. But if we go back to Asurbaniphal and his tablets of clay, the process of the author producing the document was fixed in that moment of time and the relationship between the author and the reader was contained and limited by that single document which then, because it was a tablet of clay, couldn't actually be changed. If it was a papyrus document it was rather better, you could make a palimpsest out of it. You could wipe out the old message and put a new message on it, but the relationship of space and time of production and consumption was separated by the medium. Go to the days of Guttenberg and the discovery of movable type and you then separate this relationship of space and time. Because now, the author can produce a document which can be read in many different places by many different people and each of them can change that document although they have no mechanism of communicating the change in the document either back to the original author or to all the other readers. In fact, they're unlikely even to know the network of readers who are concerned with consuming the document.
As soon as we get to the convergence of telecommunications and computing, we are into the third information revolution, because not only can we almost simultaneously engage in the production and consumption of information, it is possible for the author to know the set of all possible readers. It is possible for those readers to communicate with that author, and it is possible for all those readers to become authors on the same document with one another. So we have fundamentally changed the nature of the concept of the communication of information because of bringing together computers and telecommunications.
We have, however, a problem, because in these networks, it appears that telephone engineers have their own history, they have their own traditions, they have their own ideas and are optimising the capacity of networks. Whereas for example, broadcast engineers actually have a completely different idea: maximising the proportion of the market captured by a single message relative to that single transmission is the key dynamic. We have computer scientists who don't actually understand the tradition, either of telephone engineers, or of broadcast engineers, but who actually have completely different understandings about the amount of traffic on their processors, the degradation of their processors, access to storage capacity and so on and so forth. And then we have editors, we have people who are interested in bringing packages of information on to the market who have come from a tradition of journalism and so on and so forth. None of these 4 groups really have any tradition of understanding what one another are on about.
But basically underlying this problem of the convergence of telecommunications and computing is that we have sets of networks. The first basis of the networks that you have one- to-one communication. The second process of the network is that you have one-to-many communication where the many are known to the author. We might call that narrow casting. Then we have one-to-many communication where the recipients of the message are unknown to the author. We might call that broadcasting. Then we have many-to-one communication, but that's only an instance of lots and lots of one-to-one relationships and we have many-to-many communications but those are only in instances of either one-to-one or one-to- many communications.
Each of these 4 different types of communication actually in turn, have different economics built into them. They have different pricing mechanisms that we need to concern ourselves with. But the fundamental point, no matter the form of the network that we are looking at, is we start off by saying, that it has input devices and output devices,these might be telephones, they might be faxes, they might be work stations,they might be broadcasting stations and televisions sets. The fundamental principle is the same, we have input devices and output devices. Then we have a transmission mechanism, we have a cabling system in some sort or other; its either a real cabling system in terms of being proper wire and optical fibre,or it's a virtual cabling system, taking place as a result of radio transmission of one form or another, and we have some mechanism of switches. So the primitives of any of these networks is they consist simply of input/output devices, they consist of transmission channels and they consist of switches.
Now if we look at the concept of the network in this sort of way, then we see that we have British Telecom with a tradition of copper wires, telephones and exchanges. We have Cable and Wireless, with I think, much the same tradition. We have British Rail with a network of railway lines of signals and of railway stations where the initiator of the message then has to make the journey to the railway station but you are still engaging in the same sort of principle of a network. You have the Post Office, which has the mechanism of addressing the idea of a system of addresses where space is organised in order for a communication to take place. And then they have post boxes, they have sorting offices, but the network functions in the same way. We have the BBC, ITV with transmitters, with television stations, radio, television sets, wavelengths and so on and so forth. And we have something like, for example, the CEGB which has a network of copper wires linked to eventually almost every home in the country. I think it is still true that telephone have only entered something like 80% of the potential domestic market in Britain whereas electricity has very close to 100% penetration. And I'm sure we can think of lots of others.
So there are lots and lots of different companies; there are lots and lots of different organisations and people who are involved in this process of putting in place networks whereby communication can take place between the authors and the readers of a variety of message. The issues that have come up over the last few years are probably the issues of privatisation, the issues of liberalisation of the market, the issue of deregulation. And I'm sure you are all so steeped in those issues that I don't need to touch on them. So the network then, means I think that we've really got to broaden out our ideas about who is capable of providing the sort of network that might be interesting if we accept that my argument about the nature of the commodity called information, is as it was when I outlined it at the beginning.
Thirdly, the processors is a relatively more simple issue to deal with because I think we understand that more clearly. It's basically boxes, wires and chips. It goes more traditionally with manufacturing as we understand it. We understand something about vertical markets. We understand something about market niches. We know whether we are trying to produce the chips and mine the copper. We know whether we are interested in taking one ton of copper in at one end of the factory, breaking it down into thousands upon thousands of tiny little things called chips and then building it up again in order to produce boxes out the other end that we can somehow or other sell. I think we know something about the functioning of that market; I think we know something about investment; we know something about production; we know something about amortisation rates; we know something about depreciation and devaluation. This process we probably understand a little bit more clearly. I'm not quite so clear why it is that when everybody seems to be involved in the GEC deal, I've not actually seen any reference to British Telecom yet, but I'm sure there is a very clear reason for that, that somebody will explain to me very patiently in the bar afterwards. The fact that from today, AT&T is involved in the deal,I'd have thought might just have changed the rules that you thought were in play last night, but that you can explain to me in the bar.
The last one is the issue of services, and that seems to involve firstly, the area that we might understand most clearly, the production of software as a mechanism whereby the processors are actually turned into processors where they actually become capable of doing things. Companies that come to mind immediately are, for example, Logica.
We then move on to the area of er, well to get a bigger hiss, anyway we'll turn it into a Victorian Cavalcade any moment now where we can touch all the goodies and the baddies and get hisses and boos and what have you, because we move onto the consultancies and the ones that immediately occur to mind here are the Arthurs. Right. Why they're all called Arthur, I don't know. But clearly there is a major industry there and the bulk of them, Arthur Little, Arthur Young, Arthur Anderson and what have you, were historically predominantly accountancy and auditing companies. They have fantastic problems because their accountancy and auditing wings are not producing a profit, not producing a rate of return in comparison to their management consultancy wings. Now that's a contradiction, that's a problem for them, but it does mean that they're having to move into expanding markets and so we find that fairly major parts of software development are actually being undertaken by these consultancy companies.
I would add on to that, of course, the question of education and libraries, because if we are going to be talking about an information economy,then it seems to me that the issue of education is actually going to be a fairly major one because these primitives that I'm talking about are going to require the capacity of children to understand the world in a rather different way from which they have done in the past.
One of the research projects I did in the early 70s, was to study what children in a secondary school in East London actually need to know to be able to be urban literate and how that's relating to what they actually got in their education system. And in fact I was fairly horrified to discover that 40% of the 16 year olds had never actually been on the underground and that for the 14 year olds, if you drew a line between Dalston Junction and the school, 90%of them had not been more than half a mile outside that line. And we then went on and discovered that in fact substantial numbers of them couldn't use a Telephone Directory, that a large proportion of them didn't know how to use an A-Z. [For those of you who are not London based, this must sound terribly provincial and detailed, but in London things like A-Zs are actually fantastically important.] And if you can't use them, then my argument would be that you are functionally illiterate. Even if you drive a company Volvo, actually you are functionally illiterate. So my argument here, is that education is another area of this service and of course, I've got to put in there also, the process of designing information systems because that seems tome to be able to be the process of what's linking these things together.
So, what I'm suggesting is that this combination of information networks,processors and systems is producing for us a new sort of information economy in which a company like British Telecom is really going to have to try and produce some sort of position about what it's going to try and do in that market.
Now I've read the Butler-Cox Report on European Telecommunications Markets until 1992. I've read several of your strategy documents and I must admit that I'm not any more clear now than I was before I read them about precisely what sort of area it is we are going to go for. And in order to try and elaborate this a little, I want to draw a little on some work that I was involved in, in the city of Kawasaki in Japan, who used to make motor-bikes and because of the nasty little yellow chappies, who as far as they are concerned, are the Koreans, they don't make ships any more, they don't make motorcars any more,they don't make motor- bikes any more, they don't make steel any more, they don't make aluminium any more. All that fantastic industrial development that took place after the 1950s has been wiped out and the city government of Kawasaki opened up (and this was a fascinating experiment), they opened up a competition for information strategies for the city of Kawasaki. Now, I was at the conference at which the result was announced, and it's not an accident this competition was won by an architect, because I don't know that we have any tradition of opening up information strategies to competition, in the way in which architects do it. The conclusion might be, either you don't end up with the Lloyds building, or you do end up with the Lloyds building, but I think the process of competition actually creates interesting ideas. And the idea that came out of this Kawasaki city study, was the idea of the intelligent city.
Now the intelligent city seems to me to be a manageable idea because people understand urban spaces in some sense as not only being spatially delimited, but people understand them as being complex in a way in which urban dwellers don't understand rural communities as being complex. I think it's a partitioning of the problem purely for the point of tonight's discussion. I think one could, with no difficulty, talk about an intelligent universe, but let's limit ourselves for the moment to an idea of an intelligent city. And the metaphor which I use to try and get this across, because people begin to grapple a bit at the edges with it, is that really the motorway is fundamentally concerned with the movement of people and things. And it's in a possible alternative to the movement of people and things is the movement of information. Or you could be moving information about the movement of people and things.
But in the generic form, if we can introduce another set of primitives, that really the measurements that we have to deal with are firstly, the relationships of space and time. Secondly, that every activity of communication has an origin and a destination; that every event in this communication has decomposition, it came from being something which it is not any more, rather like that ton of copper that entered the processor factory,that ceases to be a ton of copper in the process of becoming a few million chips, and it's in a process of recomposition when it comes out the other end as something else. And those of you skilled in some of the traditions of information systems design will recognise that that's not dissimilar to the idea of entity life histories.
Now if that really is at the primitive level, what's going on in the motorway metaphor, then we can say, the bits of metaphor that we have to deal with is, at the most elementary level, things like roads, we have things like vehicles, we have things like signals. Another thing we have is instances of journeys which have destinations, have intentions, have intelligence, but those actually really involve people who we normally call drivers.
This motorway metaphor - the motorway has to be built. That involves construction and materials. We have another part of this metaphor which involves the motorway manufacturing industry. Another part of the metaphor involves petroleum and service stations, the petroleum companies. Another part of it involves the police, the Automobile Association, ambulance services. Another part of it involves the insurance industry. Another part of it involves driving lessons, licences and the MOT. Another part of it, those of you who know the M25, involves traffic jams, which involves the idea of transportation planning which involves the idea of regulation and the State. So here we've got a little bit of a metaphor for a motorway system with it's cementing net but with all the major components of that, that involve market opportunities for each of the different players that are involved in some process or other in production and consumption.
And so we then have to go on and say, well what does the motorway metaphor look like when we move it into the field of information and networks. And I think we have to first of all say, that there are receivers and transmitters. We've got the idea of receivers and transmitters, telephones, television, faxes, workstations, that all these people here, you'll remember from my introduction on the question of information, they're all conscious active agents, they're all authors and readers, they're all producers of information and consumers of information. Their activities all have conscious and determinancy about them, and they can be at home, they can be at work, they can be on the move, they're at it anytime and in many places. And we are concerned then, with these receivers and transmitters with their manufacturing and sale. Predominantly at the moment,we are talking about telephones, televisions, faxes, workstations and what have you. What we will be talking about in 10 years time is less easy to see. Now,that seems to me to be as clear a market as we can talk about the process of motorway construction, of motorway manufacturing.
Secondly, we are concerned with the question of channels. So one is receivers and transmitters, two is the question of channels. The first layer of these channels seems to be to the traditional hard wiring, copper wires and probably optical fibres. This is a mechanism of communication which seems tome to be fairly clear. Secondly, we probably have anywhere that there's line of sight, anywhere that you can have microwave transmission. Thirdly, there's anywhere that you could bung up a satellite. So here are 3 quite clear manufacturing opportunities to go back to the receiver and transmission. The channels are, first of all, manufacturing opportunities, the optical fibres have to be manufactured, the transmitters and receivers have to be manufactured, the satellites have to be launched, and so on and so forth.
Thirdly, we have the question of switches. Now the switches are then what is going to link together these different categories of channels, and the switches have to basically understand, firstly, where something is going. Secondly, the best route and chart. So there are your primitives, that every act, every author, every reader, every act of communication, is basically a packet that has to start off with an idea of where it's going and it has to have a mechanism of understanding the best route by which it's going to get there, for which there is some charge. And where these things are going to be really is a matter of sublime irrelevance. You can park your satellite in any available parking space, either in a geostationary orbit or a sun synchronous sort. Your line of sight can be from any building to any building provided you can get a line of sight, or from every hill to any hill.
Now I think if you look at channels in this way, then the idea of copper wires and optical fibres as being the real barrier to entry that British Telecom has been able to put in place historically becomes a bit weak, because it might well be that all somebody's got to do is be able to lease, (rather as happened at Kingston College of Further Education), lease a parking space on the tallest building in the area and bung up a mechanism there whereby within line of sight of anywhere in the region. (Certainly I read yesterday that video transmission over15 kilometres on line of sight is now absolutely feasible and in practice.)
Given that everything I read tells me that shunting video transmission down copper wires is a major impediment to the movement of video data might indicate that all sorts of people could come into this competition. I mentioned earlier on, the ones that occurred to me were bodies like British Rail and the Post Office. I don't see why it couldn't be any old property developing company. We already know that the major banks are making more money out of the leasing and management of the spaces above their branches, than they make from managing the accounts inside their branches. So the capacity to do all sorts of things with odd bits of real estate, if you can put the right consortium together, is fairly considerable.
And fourthly, I would say, remember we've got here that within this idea of the information intelligent city, we've got this idea of receivers and transmitters, of channels, of switches, the fourth one is the area of what I'm calling protocols. And the protocols that we grasp and understand already, isthe idea, for example, of addresses. We've already got the idea that the Post Office actually couldn't function if we didn't have in place a regulated and systematic notation whereby people's physical and spatially located activity can be accounted in a way which can be understood by people with virtually no training at all. In fact, the innovation of the Post Office over the last 15 years or so, the idea of the post code, actually then produces for us a digitised map data set whereby any particular event can be spatially located down to the level of hardly more than a few square metres. The post code at its lowest level of disaggregation at the moment, probably involves something like 7-10 mailing addresses. We also know that the entire post system involving 23 million addresses is available on a CD ROM, and so the capacity to put this address system in place and then be able to organise, either at the level of the post code as a digitised map data set boundary defined standardised polygon, or moving from there on to some more informal form of addressing, like, for example, street names and street numbers and so on and so forth, that really the bulk of the protocols for those are in place already. There's a bit of blurring at the edges in terms of how you actually structure and define the orders of the digitised map data sets and there are certainly some rather clever little games about how you handle the boundary spaces involving things like, quad tree analysis, but that's really simply in order to reduce the drain on the processing capacity, of being able to handle very large chunks of data.
But you've not only got that capacity in place, you've got the capacity from the minimum data set of your lowest possible spatial boundary to be able to aggregate upwards in order that at one level you end up with Mundocart, where you've got an entire map of the world which can either operate at a global level or can, with the click of the jolly old proverbial mouse,actually bounce down to any disaggregated level you like. Now the Mundocart,if I remember rightly, has only digitised 2 or 3 million spaces, so it's only really the major United States cities that are handled in this sort of way. If you click over into India, it's rather a sort of big blob with Hyderabad and Ahmedabad and Delhi and not very much else. But it's clear that the capacity is there. We already know that the Australian census is going to be published in 1991 on CD ROM? We know that the United States with their Tiger program are going to integrate the capacity to handle the whole of the United States census linked into an integrated digitised map data set which is going to be capable of generating data disaggregated to the level of either an enumeration district or a post code. And if we are talking about those sort of data sets which are predominantly concerned with polygons, it's then not very difficult to imagine, and I know there is one major British Telecom research project in this area, to then say, that nodes and channels within these networks are in turn, simply either where your telephone cables are lying,where the water board has got its pipes, where the gas board has got its pipes and so on and so forth.
So it seems to me that we need protocols that deal with the addressing facilities. We need protocols that are involved in the mapping facilities in order that the spatial and temporal relationship of information systems that I've been talking about can be put in place, and we need protocols about how the information is actually going to be communicated through this information city. It ought to be a matter of absolutely sublime indifference to the author the reader, how the data gets from point A to point B. That should be a matter which is outside of their concern. What they require, is to know where it's to go, to know when it's going to arrive there and to dispatch it. And it seems to me that there's a whole layer of activities that should be put in place in order to be able to satisfy these 2 requirements.
And another of these protocols, as I've said already, I think ought to be in education and training, because it seems to me that even if we are capable of thinking about this motorway network, we are capable of seeing the idea of an intelligent city that has these sorts of things to it. It has vehicles, it has signalling systems, it has a motorway construction, it has a police force and it has a driver/vehicle licensing centre and what have you. Then we are going to need a similar education and training if an entire population is going to be able to understand and take advantage of this market in the intelligent city where the price of moving any goods or person from point A to point B.
In other words the pricing in the distribution and exchange part of the model becomes so cheap, that we fundamentally change the relationship between the cost of production and the cost of consumption, for any particular commodity that we might be involved in. But this has the potential in turn, to really quite change the ideas in which we go about the process of marketing. But actually the idea that we are into at the moment of High Street retailers, the idea of home banking, the idea of smart cards and so on and so forth, have a capacity to really be quite fundamentally changed if we can grasp the way that this information city is potentially going to work.
Now how are parts of this going to develop? Quite clearly, it's not going to be any individual with a Masterplan. We are not going to have some Department of Transport person coming along and saying, well what we need is the M25 and because nobody wants it, will only build it to a third of its capacity that we know it's going to be, and because they, in turn, completely get wrong the fact that if you intervene in a transportation system then actually you don't simply redistribute the balance of distribution of transport, you actually change the pricing of transport. Any therefore, the consequence of that is, by improving your transport system, you make it cheaper to move a person or a thing from point A to point B; you therefore change the entire absorption and consumption of that network as a whole. Whereas, if you change the functioning of an information network, we are far from clear that it's going to have those sorts of consequences at all. In fact it would appear that, provided we are absolutely certain what the technical issues are, in terms of the capacity of the switches, the capacities of the channels, then actually we could change the costs of transportation of goods, services and people in ways that we really haven't grasped so far.
So, how are parts of this going to develop? Workstations; we've got a good deal of that there already. People are now using their Mackintosh environment in a way which is really quite different (and there's no hissing on that). The next one you could hiss to your heart's content. Whereas, people are still using MSDOS on those dreadful horrible little machines manufactured by a company which is unmentionable, in order to really totally cripple their lives. I don't know whether any of you saw the programme last night on television, the Goddess and the Computer, about the way in which water management systems were used in Indonesia with an environment that really people could grapple with immediately. Now the information might be wrong, the politics might be wrong, but people were able to grasp the idea of modelling extremely quickly and able to test their received models of the world on the basis of their thousand years of activity, against the models that had been produced to analyse the developments of complexity by increase in the population flow.
So it seems to me that in the foreseeable future, that sort of environment is going to be one where a very large number of people can actually engage in manipulating information, thereafter, we'll get into talking to the little machines, but that's going to be probably a few years. So for the foreseeable future, my guess is that something like the Mackintosh environment at the price that Alan Sugar can get the things into the shops, will give us the mechanism where workstations really become about as disposable as most of the other hi-fi equipment.
Possibly another part of this will come through ISDN. I think it's perfectly conceivable that core and spine wiring through involving some sort or other,optical fibres, some sort of network, large parts of that are in place already. How it's going to develop is less clear, but it seems to me that there might in turn be quite a large number of cores and spines that have got some sort of interfaces to one another. And then we've got the mechanism of being able to communicate amongst them.
A third area which might be interesting, is the area involved around Centrex. It might be that there's a capacity there by really putting intelligence into switches in order that it doesn't actually matter whether you network nodes are on the roofs of buildings and the buildings are wired. There's already quite a substantial literature in intelligent buildings and that the capacity is then for a Centrex type intelligent exchange to be sitting on a building or in a building or to go from buildings to another building which is called an exchange which is dedicated to that purpose. It might be that these buildings begin to work like token rings; that they've simply got enough intelligence built into them; that they receive a bundle of messages and they simply forward them on by the optimal route to their destination. But this idea of being able to put intelligence into switches in order that switches can be gathered into exchanges which can be spatially located almost anywhere, but as a consequence of the intelligence contained in the switches, a message can be communicated from any point to any point, possibly independently of hard wiring or alternatively by a variation in pricing according to investment capital,development capital, rate of return on capital employed; you can go through a process whereby you can optimise the utilisation of your networks to the best of your convenience.
Another part that clearly is going to be required is this idea of regulation. I'm always amused, there's a famous quote by a British Telecom Director, who said, "what we require is competition and co-operation". And I had another version of this at the Financial Times Conference on electronic financial services, where one of the bankers stood up and said, "what we want is competition in the High Street and we want co-operation in the back rooms". And what the banker didn't understand is when he talks about co-operation in the back rooms, what he means is all these people fighting with one another over who's going to end up setting standards, over who's going to control significant chunks of the market.
But the other side of the incomprehensibility of the 2 of them, this was a classic example of the unspeakable in pursuit of the inedible, the telecommunications and computing people, they were saying that what they require is the ability to compete for markets over standards but what they require is a stable money supply. Because of course, they didn't understand that what the bankers regards as competition,is being able to fight over money supply. Being able to fight over the values of currencies and the values of prices. So here we had an example where two really mutually incomprehensive groups of people were talking to one another about how these networks are not going to come into place.
Part of what we can see also though, is the development of value added data services. Bits of them are in place. They started in a large part with libraries and librarianship, Dialogue, ESA, IRS, Pergamon and so on and so forth. Firstly concerned with access to very large bibliographic data services. It wasn't long after that,that companies began to trade in financial information and out of things like Reuters and then Seac and Topic and what have you. You've got to trade in financial service which took rather a battering in October 1987 and hasn't recovered, but what it has, I think, indicated, is there are substantial other niches where trading in this information creates a potential, I'm sure that you already know about things like Trader Net, you're probably involved in Swift and so on and so forth. I've certainly seen one of your documents which identified half a dozen market niches where you thought that value added data services was where you ought to be in place. So bits of this information city are in place already. Rather like, you remember the film, is it The Life of Brian, you know, what did the Romans ever give us, gave us aqueduct here,but apart from aqueduct what did the Romans ever give us. Well, they gave us housing, yes, well apart from housing what have the Romans ever given us, they gave us security and banks and so on and so forth, so we've got bits of it in place, but it's clear that we don't yet have anything that people can grasp hold of with the metaphor, with the clarity of the idea as when we talk about a motorway system. And if somebody or a company is looking where they are being located, then people actually can't grasp their market opportunities with a clarity that they can say, there's the Automobile Association, or there's the driver/vehicle licensing centre or there Austin's - actually Austin Rovers are a very bad example (despite the fact that Graham Day has just got a knighthood out of it), because who could say what market Austin Rover's into. So there are clearly major barriers in turn. Having identified that we've got bits of it in place already, it seems to me that there are major barriers.
The first is really something which is structural to the nature of capital itself. It's a bit like the Spirit of Free Enterprise sailing out of Zeebrugge with it's bows wide open. There's an enormous amount of water,enormous amount of capital, slushing around inside this tanker, but companies are not prepared to invest substantially because the rates of return on capital employed are not high enough.
Secondly there is the issue of pricing and charging. Much of this I've covered already.
What relates the organic composition of capital to pricing and charging, is what I call the proof problem in information systems design. Although we've got an enormous amount of detail in things like structured systems and data analysis and data based sizing and so on and so forth, we are much weaker in having any idea about how to relate an information strategy to the actual design of that strategy, the design of the architecture that's going to implement that strategy. And we are much weaker in being able to evaluate investment in information systems design and investment in information technology in order to justify that investment against other possible types of investment or investment in other areas of markets.
Fourthly, we've got this issue of Chinese walls which seems to be a necessary concomitant of this idea of competition. So in several of the functioning finance houses in the cities, they're not Chinese walls just in terms of people putting up boundaries that you can't cross, they've actually built physical walls in order to stop 2 departments talking to one another. And I think that as the system becomes more complex, well we've got free movement or relatively free movement of goods in these networks here. We are going to have increasing impediments to the free movement of information in networks, not simply because of the question of pricing and charging, but because of the question of privacy and secrecy that I started off with. That, because information functions as a commodity in a way different from other commodities, there are a whole series of issues built in there that are going to increase problems.
And fifthly, I think the barrier is going to be this absence of generic protocols, this business whereby we actually don't have the addressing system,we don't have the mapping system for this new information city. And so it's going to be difficult for information to successfully communicate from one place to another.
So there then, from a definition of what I understand the nature of information to be, leading on to the nature of what I consider networks to be, the role of processors and the role of services, I've developed an idea of a metaphor in the motorway from which we can grasp the idea of the information city. And I then indicated what I considered to be the opportunities, what actually in place partial already, went on to indicate what the barriers are. And it would be unfair if I didn't end up by saying that I ought to comment very briefly on some of the things that I think that it could be that a group like yours could be doing. I would put the bulk of my resources into cracking this question of intelligent switches. It's this mechanism of building up the intelligence whether they're in individual switchboards or they're in exchanges or they're in parallel processors, they're in neural networks, they're involving expert systems, they involve data bases, they involve hyper text; none of this is clear yet, they might involve large data stores on CD ROM interactive optical discs and so on and so forth.
The architecture of these intelligent switches I'm far from clear on yet, but if I was going to allocate my resources, I would say that the people who are going to control the major chunk of the market,that this information city, this intelligent city can handle, are the people who handle the switches. The idea that came out in Kawasaki was they are going to put intelligent workstations as commonly as they put letter boxes and street lights. (And a marvellous thing that's apparent which you don't get in Britain which are little coin dispensers that for a few yen throw out at you litres of beer and litres of whisky, and I've never seen one of them vandalised.) The intelligent workstation is going to be as common as the intelligent whisky dispenser. The key is going to be the switches that optimise the routing of any of those messages from their authors to their readers.
Now those intelligent switches are going to then require the second area that I think work needs to be done on, which is basically the building up of protocol libraries. But if we can go back to another old metaphor people grasp on, the idea of encyclopaedias and dictionaries. That it was an amazing shift in the 16th century to suddenly have this idea of putting information into alpha-numeric order because previously all information had been organised systematically. The idea of alphabetical order was an enormous step forward,but it then meant that you could balance alpha- numeric order and thematic order in order to be able to put intelligence into an information system and that's how a dictionary and an encyclopaedia emerged differently from things before. Dictionaries and encyclopaedias means we are going to have things like lexicons and bestiaries. We've got to have basically, maps of how these protocols are going to fit together. If we've got the lexicons, the bestiaries, the maps, then we've got to have the grammars, we've got to have the rules, the protocols, the methods, whereby these different structures fit together. Once we've got the grammars, we have then got to have the catalogues and the classification schemes in order to be able to make sense of this information, of this intelligent city. And then we've got to have the directories whereby we can navigate our way around this intelligence. And it seems to me that it's in the production of these protocols that are going to be a pre- condition to the creation of the capacity of people to navigate their way through these intelligent cities.
And finally, I'd like you to join me in helping to fight the Nannies.