|
Larry Ellison
Co-founder & CEO, Oracle
In recent months I have been publicly called "a prophet and a liar," "a genius and a fool," "dead right," "half right," and "a dork." But all I did was suggest that everyone might benefit from a radically simpler and less expensive computer.
An inexpensive and easy to use computer is essential if we want a prosperous future in America. Toward this end I will discuss the difference between pencils and PCs, and why our prospects for the 21st century are rather bleak if we cannot get our PCs to be more like pencils.
The Oxford English Dictionary tells us the word "pencil" comes from the Roman penicillium, a form of writing brush. By the I7th century, graphite stylists were being made by artisans all throughout Europe. By the late 18th century a Frenchman developed the first practical method of encasing graphite in wood.
By the 19th century the German knack for industry propelled them to domination of pencil manufacturing. Stiff competition from American factories, however, drove the price of pencils relentlessly downward. By 1900 a pencil of high quality could be obtained for just two or three cents. Combined with paper and printed books, these cheap pencils played a crucial role in the rise of literacy and education worldwide.
Most of us remember Armand Hammer as an oil baron, but he actually made his fortune in pencils. When Lenin granted Hammer the first foreign business concession in the Soviet Union, Hammer shrewdly shopped around for a profitable venture. He first thought to make Ford tractors, but Henry Ford refused to work with Bolsheviks.
Five Cent Solution
Hammer eventually decided that the Soviet Union, bent on universal education, would have a considerable need for pencils. Hammer smuggled a German pencil factory into revolutionary Moscow by breaking the machines down piece by piece. He even set up a sham factory in Berlin so the equipment manufacturers wouldn't be suspicious.
In its first year of operation the pencil plant produced 10 million pencils. By the second year it was able to cut its price to five cents, from 25 cents, and successfully export to Russia's neighbors. More plants were built, and by 1930, they made so many pencils and so much profit that the government bought him out for a princely fortune.
Pencils were the foundation of Hammer's great wealth; they were also the foundation of Russian literacy. Through compulsory education and the generous use of Hammer pencils, a country suffering from less than 30 percent literacy raised the literacy rate, within a generation, to nearly 100 percent.
What are the attributes that made the pencil so successful? It's easy to use; it's cheap; it can be used to draw almost anything, and it will process words in any language. Even a three-year-old with minimal instruction can use one. In talented hands it can create works of infinite complexity, subtlety, and beauty, and at three cents each, anyone can afford one.
Dangerous Decline
Today America is facing a crisis of education and literacy not unlike Russia's a century ago. While the number of totally illiterate Americans remains fairly low, there are a large number whose minimal literacy will bar them from success in the 21st century. Various estimates suggest that between 15 percent and 50 percent of our countrymen are functionally illiterate.
Our educational system has suffered a disaster, a tragedy of disintegration inconceivable a generation ago. It is so pervasive that we have grown used to seeing millions of marginally educated people trapped in poverty. We are too familiar with poorly schooled American workers who cannot compete with their peers abroad. We accept the relentless decline of reading and all of its consequences as a national habit.
We are now beginning to see even our literate students emerge with vast gaps in their knowledge. Many have no notion of history and cannot distinguish science from science fiction. The worst cannot find France on a map.
In some of our public schools violence and chaos has reached levels where no education is possible. We have schools that would embarrass a Third World country. Everyone agrees that we need to improve education's quality and effectiveness. The life expectancy of a modern democracy populated by the uneducated is short.
The most commonly offered solution is to provide a personal computer for every student in America; it is seen as the modern teaching machine of infinite possibilities. While this would be good for my industry, I cannot think of a plan more likely to end in disappointment.
Powerful, But Pricey
The problem is that the personal computer is no pencil. It's not easy to use; its not easy to program, and it's extremely expensive. The average computer costs more that $2500, and this does not include software, backups, updates, administration, and repairs.
It costs more than $8000 per year to equip each person in an organization with an up-to-date Windows 95 machine. At that price it would cost $400 billion every year to equip 50 million students with PCs.
The rarely told truth is that the modern PC is an unnecessarily complex device. It is, with all its various parts, nothing less than a miniature mainframe for the desktop. It was not designed for networking. It has been remodeled like an old house, and after 15 years, its showing its age.
Compared with full-size mainframes, the PC is a wonderful machine. Your dry cleaner can now handle multi-million record mailing lists. Your auto mechanic can publish repair manuals. You can run an international merchant banking business from your apartment.
But the PC is not a simple machine to use or maintain. We cannot afford to use it for mass education. The really bad news is that PC prices are not going down. Software is getting fatter faster than memory is getting cheaper.
Since the introduction of the IBM PC in 1981 the machine has gotten bigger and faster, but not cheaper. For 15 years mainstream PC's have cost $2500 to buy, and thousands more each year to maintain. This is why I believe it’s time for a new device, the network computer or NC.
Enter the NC
Think of the PC as an electronic pencil. It's radically easier to use than existing machines, because it has less in it to control. Because it has less in it, it also costs less—as little as $500 now and less than $200 or $300 within a few years. Providing 50 million of them will perhaps cost $10 billion, not $400 billion. Last week we unveiled the first prototypes.
The NC is about the size of a small textbook. It's an inch and a half thick, weighs less than two pounds, has a very powerful processor, and can be controlled by either a keyboard or a pen. What is most significant is what an NC will not have. The NC will not have a hard disk, a CD-ROM drive, or a floppy disk. It will have only one port which connects it to the Internet. You will install a network computer by plugging one wire into the wall for electrons and one wire into the wall for bits.
When you're not connected, the device functions on a couple of batteries. Instead of maintaining all your programs and documents on the hard disk of your machine, you use the network computer to access software and information stored cheaply on huge network servers. Updates and bug fixes are automatic, and administrative costs are just about zero.
Oracle is providing the operating system, the database, the access tools, and some early applications. Unlike Windows, it's an open, non-proprietary architecture. An NC can be built with any microprocessor, run on any network, and use any monitor—even your home TV.
Shipments are expected to begin in the third quarter of this year. In the future the NC will come in many flavors. There will be NCs for the office, NC laptops, and NCs integrated with phones and televisions.
When I have suggested this notion, I have encountered a level of anger and extreme reaction that makes me suspect that we're onto something. It's the kind of reaction one might of expected from the manufacturers of steam locomotives when the suggestion was first made that a somewhat smaller and less powerful device might better serve the interests of personal transportation, such as the automobile. It's exactly the same dyspeptic reaction that mainframe makers had to mini-computers, and that both of them had to PCs. They all said it would never happen.
No Dumb Terminal
I had expected some hostility from Microsoft—No monopoly, credited with an industrial miracle, wants to hear that its software is cumbersome, expensive, and doomed to a mere supporting role in the future. Microsoft claims that the network computer is nothing more than a "dumb terminal" attached to a mainframe. Microsoft claims that the PC revolution replaced mainframe terminals with PCs, and that no one will go back to relying on a network for software and data storage.
In fact, the network computer is no dumb terminal. It will have as much raw processing power as today's most powerful PCs. As the Internet develops, we will come to rely on it just as we rely on the networks that deliver our phones, power, and drinking water. Microsoft shouldn't worry too much. Just as the PC did not replace the mainframe, the NC will not completely replace the PC.
There are more than 40,000 different PC programs and applications. The network computer will only be ideal for a few applications, such as E-mail, word processing, video telephony, simple spreadsheets and presentations, and access to the vast resources and databases of the Internet.
These tasks account for 95 percent of all PC use. The NC will be so good at these important tasks, and so inexpensive, that it will soon outnumber PCs. Most important, because it will be so much simpler to use and an order of magnitude cheaper, it will bring the benefits of computing and information to the entire nation, not just to a select class.
Imagine the educational impact of computer technology that allows every student and teacher to access text, audio, and video libraries so they can communicate with each other or look up any book on their screen. Imagine the impact of compelling high quality, full-motion interactive educational video software that can compete with Hollywood and MTV for our children's attention.
My own industry has not set a shining example in this area. While the computer may be the most important educational tool since the pencil, very little effort has been spent to develop the software needed to make it so. In 1993 more than $8.5 billion was spent on home video games.
During the same year, Americans spent only $150 million on retail educational software. We all believe that education is much more important than entertainment. We must make it at least equally compelling and successful as entertainment if we want our nation to prosper. We must begin building all the tools needed to make that a reality.
Untapped Potential
Oracle, for example, has developed a universal server. Unlike other multimedia systems, which can support only a few simultaneous video channels, Oracle's can handle thousands of video channels on a very small machine that you could put in a school. It allows students to view outside material on the Internet, and it allows teachers and administrators to be selective about access.
Like a good school, it provides a safe and supervised environment, free from chaos and violence, yet is open to exploration, creativity, and real learning.
The computer's greatest benefit to society will go untapped if we cannot engineer the end user machine at a reasonable level of ease of use and cost. The emergence of the Internet gives us a new way to achieve universal access. It allows high performance computing without requiring everyone to purchase a desktop mainframe.
I believe that the dominance of the PC will prove to be a temporary, transitional phase in the information age, not unlike steam power in the industrial age. The era of the PC is almost over, and the era of the NC is about to begin. The time has come to build a modern pencil.
