e2001 Your Digital Future - Bangkok Post Special Supplement

THE CONNECTED FUTURE

E is for exponential communications growth

The Network Effect means many changes

Tony Waltham

Much of the buzz around computers and the Internet and the impact they have had on our lives can be attributed to some key underlying trends identified separately by three individuals, namely Gordon Moore, Robert Metcalf and George Gilder.

Pioneers in their respective fields, they have each highlighted highly significant trends that share a common but highly unusual mathematical attribute. And, in a very real sense, these three gentlemen have become high-tech folk heroes by having these observations attributed to them as a "law" _ so we have Moore's Law, Metcalf's Law and Gilder's Law respectively.

This attribute or quirk, if you like, is that of exponential growth _ meaning that the rate of change doubles in a fixed period of time.

The three trends govern the density of silicon chips (processing power and computer memory _ Moore's Law says it doubles every 18 months); the value of a computer network (Metcalf's Law tells us it increases at a rate of the square of the number of connections) and bandwidth (this is now tripling every 12 months, according to George Gilder).

Strictly speaking, Metcalf's law differs mathematically from the others, but the growth curve for computer networks follows a similar path of exponential growth as the increased value sucks in ever greater numbers of people.

But let me briefly explain about the phenomenon of exponential growth. This occurs when something doubles in size in a fixed period of time and it is very interesting since, when it happens _ and this is relatively rare outside the world of nature and biological reproduction _ the impact can be explosive.

To illustrate this, let us suppose that a child comes to you and offers to work for one satang a day for one month, but that his salary is doubled each day.

That is one hundredth of one baht he earns on the first day, two hundredths the second day, four hundredths _ four satang _ the next, and so on. Is this any way for a kid to get rich?

Well, you might be surprised and it would have been a way for you to be quickly impoverished. If you had signed such a contract and hired a child on this pay scheme on New Year's Day, you would be out of pocket by over 10 million baht for just one day's work on the last day of January.

Now people tend to remember this bit. But what they often do not realise is that, using the same scheme, the boy would only have earned a total of 1.3 million baht if January had been three days shorter (that is to say, if the same scheme were to have been started at the beginning of February).

To put this another way, the total income for that whole month of February with just 28 days would be roughly 2.6 million baht, instead of the 21 million baht during January, with its 31 days. It's those last three days that make the big difference.

And, obviously, the faster that something doubles, the more surprised we are when it pops up unexpectedly on our "radar screens".

It is these factors that are driving this accelerating pace of change and which has made it possible for computer experiences to become richer and richer.

Microsoft Corporation is credited with the following metaphor: "If the automobile and aerospace technology had exploded at the same pace as computer and information technology, a new car would cost about $2, and go 600 miles on a thimble of gas. And you could buy a Boeing 747 for the cost of a pizza."

When I first accessed the Internet from Thailand back in 1992 (before universities here were connected permanently to the outside world), it was at 2,400 bits per second and I used my computer as a dumb terminal to a character-based Unix server.

I had to remember commands like sz to send a file by a protocol called Zmodem and then I had to remember to press the Page Up key on my computer to upload a file.

Today, I access the web at 56 Kbps, use a graphical interface and typically have multiple file and data transfers taking place simultaneously without even having to worry or even think about it. Similar progress has been made in terms of processing power and storage.

And the World Wide Web has become a mainstream environment, but, always a rapidly-changing one. For while today we access web sites almost exclusively from desktop computers, in the near future we will probably use a smaller, more portable device to meet our growing needs to communicate and to access information.

And we will do this at any time we care to when this is convenient and anywhere, particularly while stuck in traffic jams or at 35,000 feet in an aircraft, two situations when we have plenty of time on our hands.

There has been a lot of excitement recently around data-enabled mobile phones, or so-called WAP phones, and it is certain that cellular phones we carry around today will morph into information tools, combining with PDAs such as the Palm Pilot to put everything into one package in our pockets.

These devices will surpass PCs as the most common way to access the Internet in the near future, while also coming soon is a whole new range of so-called Internet appliances for use in the home, at work or in the car.

They have been called Web pads and e-books, while automobile dashboards will soon display where you are and offer advice on where to dine or give shopping bargains in addition to telling you how to get to a destination.

It has even been said that as you begin to get low on gas, the car will automatically open an auction and filling stations in the area will bid for your business _ with the on-board computer accepting the best offers and possibly letting you, the driver, decide between the most convenient or the most economical tank refill. Once you decide on this, it will tell you where to turn off.

Meanwhile, heads-up displays using goggles are being modelled on cat walks today and the fashion accessories, even clothing of tomorrow will be smart _ and networked. There is even talk of active contact lenses that in a few years will allow you to view fully realistic scenes literally without opening your eyes.

All of this and much more is the almost inevitable outcome from the exponential rate at which microelectronics is packing more processing power, more storage potential along with other features including sensors and video lenses into ever-shrinking packages.

Then there are the technologies and standards that will be the glue that will stitch the fabric of our networked future together.

These include Bluetooth that will govern the wireless links around us in the home and in the office, JINI for the way that devices will probably "discover" the presence of other active devices and learn what they can do together or to talk to each other, and GPRS or G3 for the next wave of high-speed cellular services that will bring faster Internet than we have over a PC and a dial-up modem to a mobile phone.

Coming even sooner, and possibly already here in your neighbourhood in Bangkok, is the so-called broadband access, which I will very loosely define here as a data link that is faster than a regular modem connection, for which the ceiling is around 50,000 bits per second (bps), but often slower.

Broadband starts at data speeds of around 100 Kbps and extends up to many megabits a second and these faster connections can come in various ways to just about everyone here in Thailand today, either by satellite, by cable or by ADSL modem links over your existing telephone line, through services that are already now being offered by both ISPs and network providers. There is also ISDN in this alphabet soup, but as a much older technology it is limited in the amount of bandwidth it can provide and now looks to be leapfrogged by DSL technology.

Unfortunately, the telecommunications environment in Thailand is not free or open and its highly-regulated status for the foreseeable future may inhibit progress here. We may instead have to be content to read about the success of wireless data services in Japan and Singapore and in other markets before we can fully take advantage of them here.

However, you can bet that they will arrive here eventually, although possibly at a premium _ and also because of the need for such connectivity for visiting businessmen who will be bringing their latest gadgets with them to Thailand and who need to remain in touch more or less constantly.

We are now entering a transition phase in how telecommunications systems work. We are moving from switched, dedicated (voice) circuits, to packet-switched data circuits. And voice will be just one form or subset of the data being sent, while the others will be the video, audio, images and text streams that can be found on the Web today.

And since traditionally packet-switched data or IP networking over the communications, especially for long distance conversations. Again, telecom munications deregulation should help to open up competition in this arena as well.

It has been said that IP networking will soon become so pervasive that in a few years' time it will be as important a foundation for the civilised world as electricity is today.

There are likely to be many new ser vices that will spring up around such a ubiquitous network as the Internet and they can be hard to predict because networking, as it links more and more people together, has its own set of rules that contradict conventional beliefs and encourages or rewards innovation.

Take, for example, the way that the new MP3 file format and the Napster music sharing site completely caught the music industry by surprise and while the Recording Industry Associ ation of America has frantically turned to the courts to try to assert its rights, most commentators say that the ``cat is now out of the bag'' with regard to the sharing of music files.

As mentioned at the beginning of this article, Robert Metcalf, the founder of 3Com and the inventor of Ethernet, is credited with having said that the value of a network is a function of the square of the number of its users.

To illustrate this, think of the value of the first fax machine. It is not very useful at all. However, once the second machine is sold, then there is at least someone to communicate with.

However, as more and more fax machines are sold, so the device becomes more valuable _ increasing at a rate that is out of proportion to the numerical increase.

Today, when you buy a fax machine, you are not just buying a box for a few thousand baht, you are purchasing the entire network of all the other fax machines and all the connections between them.

Stemming from this, the ``network effect'' suggests that the more plentiful things become, the more valuable they are. This notion directly challenges two older beliefs _ that value comes from scarcity (oil or gold) or that the more plentiful something is, it becomes ``cheapened'' and the less value it has.

Networking brings its own set of economic benefits, especially where the product is digital _ where copies effec tively cost nothing to make. If we look at a chart of Microsoft's soaring profits over time, it reveals that during its first 10 years, its profits were negligible. They only rose above the ``background noise'' around 1985. But once they did, they exploded.

Federal Express experienced a similar trajectory: years of minuscule profit increases, slowly ramping up to an invisible threshold, and then surging skyward in a blast sometime during the early 1980s.

Similarly, fax machines had a slow growth for some 20 years and they were suddenly everywhere in the 1980s, while the Internet is another good example of spectacular growth.

In this so-called network economy, we also have what is called ``the law of increasing returns,'' which is revenue that is created and then shared by the entire network _ witness how Micro soft's own success has been shared by tens of thousands of companies in the business developing hardware, soft ware and related products.

Other network economy dynamics are that prices continue to fall, approaching the free, down to and including free for some products. Ready examples here are the Web browsers from Netscape and Microsoft, RealAudio streaming software or cellu lar phones in the US, where they are given away when you subscribe so that you simply pay for the service as you use them. Nor can we ignore the way in which the Internet explosion has also caught the imagination of technologists and commercial interests alike and it has been the combination of investors and entrepreneurs along with programmers and product developers that has been the driving force behind the rapid inno vations around the Internet that we have seen over the past five years.

For the Internet, hypertext mark-up language (HTML), the TCP/IP protocol and Internet-oriented languages such as Java, XML and now UDDI, which stands for Universal Description, Dis covery and Integration, are serving as the all-important building blocks for future networks.

The network effect is providing prod ucts and services that are essentially free _ after all, how many web sites do you know for which you need to pay a sub scription fee in order for you to access? And we are already seeing free PCs and other Web access devices being offered to those who subscribe to a service, while the cost of Internet access has steadily fallen and a free Internet service provider is now ready to start up here.

In the network economy, ``free'' goods tend to have links to enhanced products or services that are or can be paid for, and this model applies particu larly where copies cost nothing to make, such as with computer software.

Many software firms offer a free ``light'' version of an application, and a full or professional version that a user can upgrade to and pay for if they wish.

In other cases, software is free to indi viduals for private use, while businesses or companies are expected to register and pay.

In yet another ``free'' model, Internet software such as the Eudora Pro email client is provided free because it comes with advertising, or you have an option to pay for the same package for the privilege of not having the advertise ments.

Since cheaper is better and, in this network economy, the more wide spread a product becomes, the more valuable it is, so it follows then that the best price is free.

Indeed, what is offered for free today may well be of value tomorrow. While Hotmail, the Eudora email client, many Internet guides, search engines, FAQs, remote live cameras and most online publications are free to use or to access, each of these popular products or ser vices have or will most likely have prof itable companies built around them.

However, while the Internet is enabling an entire new world of interconnections, another key element in the high-tech world of today is the ability to build smaller and more power ful devices to store, manipulate and display this digital data we can connect to.

The good news here is that Moore's Law still has a lot of life in it. This is the law that is named after Intel Corpor ation co-founder Gordon Moore and follows his pronouncement made in the 1970s when he predicted simply that every 18 months semiconductor density would double _ meaning faster and cheaper processors and greater mem ory storage.

It used to be that Moore's Law would run out of steam in around 10 years time when atomic limits would be reached.

However, this has since been revised and Sun Microsystems Chief Scientist Bill Joy explained in June this year that breakthroughs in molecular technology would now enable us to continue to meet or to even exceed Moore's Law's rate of progress for another 30 years, or until around 2030.

By that time we would be likely to be able to build machines in quantity that will be a million times as powerful as the personal computers of today, he said _ and he is more than a little concerned about the possible ramifications of the power of this technology _ but that's another story.

Another area where a similar expon ential growth is occurring is in network bandwidth. It is Gilder's Law that governs bandwidth, named after George Gilder, a radical techno-theorist who forecasts that for the foreseeable future or over the next 25 years the total bandwidth of communication systems will triple every 12 months.

His argument is that, in essence, bandwidth will become free. And he points to radical developments in optical technology _ where essentially how each colour can provide a separate, wide channel for data _ that will allow this to happen.

In an interview in .ifForbes Magazine .nfin August Mr Gilder said that since the commercial advent of wave division multiplexing in 1996, ``bandwidth has been increasing at least four times as fast as Moore's Law, if not faster, and promises to continue to do so for at least the next several years.'' There is another dimension to data communications, and this is the ``other'' next big thing _ wireless. To take the words from an Agilent adver tisement in the July/August .ifTechnology Review.nf: ``It took 100 years to connect the first billion people. The second will take only five. The wireless revolution is at hand.'' But we don't need to rely on advertis ing copywriters in order to be optimistic about the growth of wireless. Goldman Sachs says that by 2003 there will be one billion wireless phones, while IDC pre dicts that in just 15 months, more Amer icans will own cell phones with Internet access than Net-connected PCs.

The first broadband or 3G wireless networks are now poised to be rolled out, offering 386 Kbps mobile links and 2 Mbps fixed links, and if you are looking for an indication of what this might mean, Texas Instruments Chief Execu tive Officer Tom Engibous gave a fore taste at a seminar in Tokyo earlier this year.

He said that wireless Internet devices would not only capture some existing PC applications ``but would introduce brand-new applications that the desk- top PC has no way to handle today.'' ``I think the availability of a wireless device that is online all the time with brand-new applications that the desk- top PC has no way to handle today.'' ``I think the availability of a wireless device that is online all the time with broadband data capability ... offers the possibility of applications that Silicon Valley is just beginning to dream about,'' the TI CEO added.

If we look at computers today, they are fairly clearly identifiable objects, usually a desktop unit with display and keyboard, but this is already changing.

In one sense, the computer itself is shrinking, and we are seeing more devices like the shirt pocket-sized Palm Pilot around us.

And then, coming from another direction, are other devices that have a computer inside them, such as cellular phones.

The definition of a computer will become increasingly blurred. And as it is tending to be primarily a device used to access and send information over the Internet _ also known as an Internet appliance _ so the image of what a computer is and is used for will change.

If we look closer at the cellular phone, these devices really are a form of net work computer _ when we switch them on, they look to the network. And they appear poised to team up with mobile computers or PDAs in the very near future.

Indeed, they may well become the mobile computer or palmtop device that we carry around.

To quote George Gilder again: ``The most common PC of the Telecosm will be a digital cellular phone.'' Nor can we overlook the fact that Palm Computing and Motorola are now working together on a combination phone/PDA.

So far, Mr Gilder's predictions have been remarkably accurate. For the future of the cellular phone, the Feb ruary 1999 edition of the .ifGilder Techno logy Report.nf says: ``The cellphone will be as mobile as your watch and as personal as your wallet. It will recognise speech, and it will navigate streets, and it will collect your mail and your pay cheque, and help you spend it.

``It will command an Internet address and a Java runtime engine, and link to a variety of displays, keyboards, and other input-output gear through radio frequencies and infrared pulses.'' Two years ago, digital mobile phones surpassed PCs in the number of units shipped and soon they will exceed the number of fixed line telephones.

``The Internet is the killer app for wireless data,'' according to Mr Gilder, while by 2002, the market research firm Telecompetition claims that nearly 12.6 million US consumers will be spending more than $5 billion to connect to wire less networks.

Just as cellular phones and compu ters are shrinking in size and price, so they are on a path of convergence _ and the main challenges ahead would appear to be design issues in making them convenient to use or to interact with.

Just how these issues are surmounted is as much a fashion issue as it is one of technology, for manufacturers today can build just about anything _ it all depends on what the customers say they want, or, perhaps on how effective marketing campaigns are in convincing them of what they think they want! All these predictions are based simply on the fundamentals already outlined: continued improvements in processing power and in the miniaturisation of chips, the parallel increase in storage capabilities and also, most significantly, the projected increase in bandwidth, with the new applications that broad band and wireless access will deliver and make popular.

We have most of the pieces or the building blocks of the future here today _ a collection of standards, protocols and computer languages including IP, XML, Java, Jini, UDDI and Internet- driven technologies.

If we couple this with compelling business incentives and software developers who will get a personal cre ative thrill from being the architects of the future of humanity, and you have the glue that will create the extended and expanding global network or matrix of the future _ and it will happen over the coming few years.

And it is beginning now. * Tony Waltham is Editor of Database.

Email: <tony@bangkokpost.net>.