Electronic learning is to education as electronic commerce is to economic activity: a disappointment in its current state of development, but so indubitably the future that one day the prefix e will seem redundant. Every educational experience will have its technological, networked element--how much so will depend on the subject being taught.

The appeal of e-learning to educators and students can be quickly summarized: e-learning allows students to receive instruction when it is convenient for them (what is called "asynchronous instruction" in the jargon of the trade); it frees educators from the tyranny of geography; it increases the resources available to both educators and students; and it improves the efficiency of education, reducing its costs while, at the same time, permitting much more personalized instruction.

And yet, of the $2.7 billion invested in e-learning in 2000, an inordinate sum is gone. Most of the e-learning companies founded in the last three years have failed. In particular, the attempt to use the Internet to reform American education from kindergarten through 12th grade has been ruinously expensive and fruitless. One company alone, JuniorNet, burned through $100 million--with little to show for it. Only two markets--corporate training and higher learning--have seen many successful ventures.

External corporate training, which enables employees to learn vocational skills from organizations outside their own company, was a $15 billion industry in the United States in 2000, according to IDC, a market research firm. It is expected to grow to $25 billion by 2003. Institutions of higher learning, to which Americans paid $11 billion in 2000, have invested cheerfully in s0-called distributed learning. According to the Parthenon Group, a private-equity firm in Boston, 10 percent of the time spent on corporate learning and higher education in 2000 involved electronic education; the firm expects that figure to rise to 18 percent by 2002. Such hopes are not unreasonable.

What's up? K-12 e-learning failed because America's public schools are run by an unimaginative, conservative bureaucracy opposed to innovations that will disrupt their settled practices; because capital purchasing in those schools is characterized by long sales cycles; and because U.S. schools do not have much money to spend. None of these factors encouraged the adoption of capital-intensive, revolutionary technologies. Most crucially, low prices contributed to educational "content" of modest quality.

But e-learning in corporations and academia works well because such institutions benefit most from what electronic education has to offer. In addition, and not incidentally, companies and universities have real buyers, with large endowments and budgets, who are committed to innovations in training and education. The higher prices that companies and universities are able to pay has justified better, more expensive educational content.

What is most striking is that e-learning has succeeded in corporations and universities despite significant obstacles. E-learning lacks common technological standards and a complete infrastructure for the delivery of material across networks; the design of e-learning classes, with very few exceptions, has been disappointing; and, most importantly, the promoters of e-learning have been curiously hesitant to create a "blended experience," combining technology-based training and human instructors. But as these problems are solved, e-learning will prosper where it has already established a beachhead. And as the cost of the technology drops (as it will), K-12 e-learning vendors will offer better material, and e-learning will become part of children's education, too.

The failure of e-learning's promoters to create blended classes derives, in part, from a kind of absolutism about technology, but more because they knew that most existing networks could not support streaming interactive video. Ironically enough, it will be the development of classes that blend living educators with

technological tricks that will most benefit e-learning. This is because not all subjects are equally suited to the course work that characterizes e-learning. Subjects in which there is a clear answer to most questions--like accountancy or Cisco-router maintenance--require less in the way of traditional pedagogic tools like tutorial and lecture than, say, poetry or political science. "It's certain that for maximum educational efficacy, there has to be a blended solution of a human instructor and a networked technology," notes Tony Tjan, the vice chairman of the advisory board at the Parthenon Group.

Blended classes will deconstruct the very notion of e-learning. One day, almost all education will have some element that partakes of information processing and communications--in common with most of our future experiences.

Write to jason@redherring.com.