At its unveiling last March, the Sun Microsystems computer language named Java was thought to be a curiosity. Its creator, James Gosling, was quoted as saying that he didn't really know what might become of it.
But by May, The New York Times wrote that by using Java, programmers could "... leapfrog the walls that now keep many computers from interacting easily with each other." Now, this "curiosity" is changing the way banks will deliver information to their customers.
"Java's promise is that it will permit applications to run across multiple platforms and networks, and that the applications will also be reusable," said Dan Schutzer, director of advanced technology for Citibank in New York, and president of Financial Services Technology Consortium, a North American group of about 80 financial services companies that includes some 20 banks. Java is already changing the way banks think about how they should deliver their products.
"Java is a specific language that facilitates the development of what we call 'distributed objects' that can operate on different computing platforms," said Dave Sexton, an executive vice president and the chief information officer for State Street Bank and Trust Co., Boston.
Java can be used to deliver text information over the Internet. Its sister language, Hot Java, can deliver three dimensional graphics on a World Wide Web page. It can also allow users to interact with the servers providing the information. The promise of Java is a new paradigm in information delivery. "But this is still research," Mr. Schutzer cautioned. "This is the promise."
About 30 years ago, banks began using mainframe computers from which they downloaded information to customers. It was 15 years ago that customers began to use desktop computers. As the computers became more powerful, they allowed customers to store more data and larger programs, which "empowered" them. But empowered customers are expensive for banks, because the customers cannot manipulate the information they have as quickly as their banks can.
Thus, banks have been forced into spending more time and money installing and servicing programs for their customers. Java can improve, make cheaper, and broadly distribute their information via the Internet.
Instead of installing big programs in the customer's operating system, Java will permit specific applications to be downloaded to customer PCs using any language. Java applications compile, as computer programmers say, to an intermediate code that can be read by SPARC, OS/2 (with PMX extensions), all versions of Windows, and Unix. To use it, all programmers have to do is write a program called an applet, which is an interpreter for any platform. It means that banks can now begin to use the Internet as a distribution channel for their products.
This will have five effects. First, banks will be able to get information into a client's hands more easily and cheaply than ever before. Mr. Schutzer noted that, "People will be able to work out of their own environment which has enough common standards and dropability so that they can almost seamlessly bring down a Java applet."
Second, banks will reduce their product distribution costs. In some ways, banks already have. "We don't send people out to install software now. It travels via diskettes across or on our network," Mr. Sexton said. But Java's promise is that it will reach a wider audience because of the interchangeability feature of its applets. "You won't even have to notify people," Mr. Schutzer explained. "They will get the update (version) when they run the program."
Mr. Schutzer said he believes the third effect is that banks will increase the life cycles of their products because they will be able to break a big program down into little applets. If, for example, a word processing company wants to update its dictionary, customers can simply download an update from the Internet. Java applets can enter any software architecture customers might have that would save the distribution expenses of putting out a new version on disks. Updates could be made when needed.
Mr. Sexton said he does not necessarily agree. "Things change so fast that I am reluctant to state this development will increase our products' life cycles," he said. "It's simply too soon to tell."
The fourth effect is that Java will allow customers to have an interactive electronic relationship with their banks. "Keep in mind that the big pools of pension monies have thousands of participants who are usually individuals," Mr. Sexton noted. Java will permit a higher level of information delivery to plans' participants.
Finally, banks will engage in mass customizations of their products because customers will be able to tap into a library of applets using only what they need.
"Where you make money delivering information is by sending data to clients or customers in highly customized applications," said Eli Kuslansky, a partner at Unified Field, a New York computer design company that makes customized interfaces. There is now a tremendous cost in distributing them that Java can cut significantly.
"For example, a bank could very easily send a 3-D applet that would let customers look at their portfolios, change interest rates or marketplace assumptions to posit what your exposure and performance might be like under different scenarios," he suggested.
But there are impediments to the new paradigm inspired by Java. There is not enough bandwidth in some parts of the world - the capacity for telephone lines to carry large programs over the Internet. There are disputes about intellectual property rights - who owns what patents to which technology. There are encryption concerns about what methods should be used to safeguard data.
"There is a claim that Java is (inherently) secure," Mr. Schutzer noted. This is because the language authenticates itself. The idea is that Java's remote programming all but eliminates the need for ongoing communication between a customer and a bank's server because the applet is a "package" that stands alone in a bank's server that has packaged instructions that activate it. The customer needs a code. The applet contains code identifying it as a legitimate source of the information. Thus, contamination by viruses is automatically limited. But this claim will have to be well tested before it is accepted.
There is also the question of finality. How can a bank be assured that a Web page transaction done with a customer will be accepted as final? "These are complex issues," Mr. Sexton said, "and I don't think anybody has got solutions right now."
There is also the question of whether to use the Internet or not. "A stand-alone system (meaning a private network) is a dinosaur," said Paul Saffo, who for the last 11 years has specialized in the long-term impact of technology for the Institute for the Future in Menlo Park, Calif. "For the next two years, if you are not putting your information on a Web page, you're not happening."
Customers want to be able to do all their dealing from one platform. "Like having one telephone to call anyone, people will want to use the same network for everything they do on their home or office computers," Citibank's Mr. Schutzer said.
The final impediment is whether Java will end up being the interactive language of choice. No one yet knows, but with Java, Sun Microsystems has obeyed "The Law of Leadership," the first of the 22 Immutable Rules of Marketing in a book by Al Ries and Jack Trout. The law said, "It's better to be first than it is to be better."
Lindbergh was the first to fly the Atlantic solo. Bert Hinkler, arguably a better pilot, was the second. Who remembers Mr. Hinkler?
Point taken, but however it turns out, the Java-inspired paradigm shift in delivering information will improve how products are delivered, and save time and money. One way or another, Java should be good for banks.