(FORTUNE Magazine) – The end is near. Yea, I say unto you, we face apocalypse here in Silicon Valley. Consider this, my brothers and sisters: Whence will our food and our sustenance come when we attain the Holy Grail, when we have indeed automated the last jot of our joint enterprise?

I found my mind wandering a bit during a particularly boring speech about software agents at my fellow columnist Dick Shaffer's latest conference, Enterprise Outlook. I began to have a doomsday vision: What if we automate 100% of the activities of large corporations? Where will the computer industry go next?

At this conference, Shaffer managed to get virtually all of the venture-backed companies that are developing information technology for big corporations into one hotel to give their best investment pitch, each in 20 minutes or less. For a venture capitalist or investment banker, Enterprise Outlook is an incredibly efficient way to get perspective on developments in enterprise automation. Since I tend to focus on software for the enterprise, four companies in my personal portfolio and another three from our firm's portfolio were among the 100 or so that made a pitch.

The theme that ran throughout the conference was automating the stuff inside companies that couldn't be automated with the previous generation of technology because it would have been too expensive. The new tech companies, many of which are just getting ready to deliver products, talked about the automation of information distribution, purchasing management, and marketing campaigns, and even about ways to offer customers do-it-yourself service online. It was this notion of automating the last corners of the enterprise that set off my dark vision about the end of corporate computing.

Let me give you a (very) brief history of enterprise computing. Computers were invented around 1940 and were first used by companies in the 1950s. Since then, computer automation, lately known as enterprise resource planning (ERP), has gone through three phases: Big Computing, which was the original vision and drove the computer industry into the 1980s; Personal Computing, which took over in the 1980s and drove the industry until the year before last; and Networked Computing, which is driving the industry now.

Big Computing involved large, complicated machines--mainframes and then minicomputers--that were installed in special rooms and managed by special people. Normal people didn't really understand how they worked, but companies succeeded in automating maybe 30% of their activities with big computers, mainly financial and operational tasks like the general ledger, manufacturing and distribution, and payroll. (Accepting a statistic like this requires a certain suspension of disbelief on the part of my dear readers, so give me some latitude, okay?)

Personal Computing brought a shift to little computers that sat on our desktops and were usable, just barely, by us. The computer industry figured out how to connect PCs into larger systems using local-area networks. This led to the development of client-server computing, which lets people in companies do something useful with only a few computers connected to each other, and at a much lower cost than using even a small minicomputer. Let's say that client-server computing brought the level of corporate automation up to about 60%.

But client-server technology had an unforeseen gotcha: To accomplish anything in concert, computers had to have common software and a standard way of talking with one another. That was fine for two computers or even four or five or ten, but became horrendously complex when you wanted to have 1,000 or 10,000 or 100,000 computers talking with one another.

Then the Internet and World Wide Web came along and solved the problem. With a Web browser, you can exchange information with other computers without installing new software; indeed, the Internet provides an absolutely standardized way for any computer (even including all those old mainframes) to talk to any other. This made it incredibly cheap for the industry to develop new ways of using computers inside companies.

So now we are in the age of Networked Computing, where a company can go all the way. Companies give customers access to their own records, give employees control over their own benefits, let financial departments control purchases of office supplies in innovative ways, let marketing departments open the development of ad campaigns to others in the company.

This is pretty exciting stuff. When you want a new computer, for instance, you can log on to your intranet, see what the company will let you buy, check out reviews of the available models, click on your favorite, and know that the corporate computer system will move everything along--getting approval from your boss and the IT department, placing the order with your company's supplier, and letting you check its progress without calling the IT department and having some nerd treat you like scum for wanting to know when you'll get your new machine. You can use the same system to get new business cards or pens or whatever, without having to send your assistant to Office Depot with petty cash.

Or you can check on the allocation of your 401(k) retirement account and discover you've been favoring public equities a bit too much and could be jeopardizing the nest egg you've been counting on for retirement. The program might recommend that you reallocate some funds to more conservative investments and let you make that change right away.

Anyway, the visions and opportunities are plentiful, as the companies that made presentations at Enterprise Outlook described. But being the jolly fellow I usually am, I wondered what happens now. If you can sit at your desk or in your hotel room or airplane seat and take care of all the big and little things that are important in your job (apart from actually interacting with fellow employees), what the heck is Silicon Valley going to do for its next act?

You've got to understand that we can talk about home computing or the information highway or computers in education until we're blue in the face. The point is that at least 60% of the hundreds of billions of dollars in annual revenues collected by the computer industry come from companies with more than 100 employees. This is the economic engine that has given Intel and Microsoft the rich monopolies of such interest to the federal government, that has let IBM and Hewlett-Packard and Compaq each amass sales of more than $37 billion a year, that has brought on such an explosion of venture capital that it no longer is considered risky to invest in startups, that has made Silicon Valley such a dense concentration of wealth that it's hard to drive faster than 20 miles an hour on a highway because of all the Mercedes, Lexuses, and BMWs. If the enterprise computing market dries up, how many of them will be rusted hulks?

I promise I'll be more upbeat next time.

STEWART ALSOP is a partner with New Enterprise Associates, a venture capital firm. Except as noted, neither he nor his partnership has a financial interest in the companies mentioned. Alsop may be reached at stewart_alsop@fortunemail.com