Don't bet on it!

Conspiracy theories run rampant in the high-technology industry, and for good reason. The unexpected often occurs.

Ten years ago, after all, few predicted that a pork-rind-eating grad student at the University of Illinois would revolutionize world commerce. Finland to emerge as a leader in wireless communications and operating systems? The decade of Belgium seemed more likely.

Still, there is a collection of commonly heard predictions that will likely never come true. Videoconferencing, micro-payments and electronic books were expected to be big at one time. Had the general public not found them irritating, they might have been.

Nothing is impossible. In early 2001, Webb McKinney of Hewlett-Packard said PC mergers don't generally work, while Compaq Computer's Mike Winkler and others said Compaq would endure. Five months later, HP bought Compaq. Nonetheless, here's a crop of things that may never come to pass.

1. Apple Computer will adopt Intel chips.

A deal to bring cheap, megahertz-crazy Intel processors to the Mac almost happened in the 1980s, said former Apple CEO John Sculley. "It was probably one of the biggest strategic mistakes Apple ever made," Sculley said recently.

The problem? If Apple ported its operating system to Intel and sold the OS in stores, a clone industry would spring up instantly. Even if lawsuits clamped down on manufacturers, consumers would make their own. Apple would find itself besieged with cheap knockoffs.

2. Microsoft will move to Canada.

Under this theory, Canadians would provide a safe haven for Microsoft from U.S. regulators in exchange for jobs.

Two problems. One, it's based on the notion that Canadians just moved out of log cabins and will do anything for money. That definition actually better describes my home state of Nevada. Second, there is the matter of export laws. Microsoft has to avail itself to U.S. jurisdiction to get its products sold legitimately here--that's why the EU can investigate. Moving headquarters won't help.

3. Microsoft is actively engaged in evil.

True, the Microsoft corporate campus is sort of shaped like a pentagram.

True, the Microsoft corporate campus is sort of shaped like a pentagram. But the more logical explanation is that they have too many hyperaggressive personalities working in Redmond, Wash. The company is also clever at contract writing and negotiation.

Still, I do like the image of Bill Gates running around the headquarters shouting, "Raikes, Ballmer, get the cloaks and the sacrificial chickens and meet me in the ceremony room."

4. IBM will buy Advanced Micro Devices.

IBM has nearly all of the elements to become a formidable competitor to Intel: a strong chip R&D group, a lot of money, and a state-of-the-art fabrication facility in East Fishkill, N.Y. AMD, on the other hand, has a chip design--the Opteron--but it needs money and a way to get into corporate accounts.

Invite Michael Kanellos into your in-box Senior department editor Michael Kanellos scrutinizes the hardware industry in a weekly column that ranges from chips to servers and other critical business systems. Enterprise Hardware every Wednesday.

So why won't it happen? Competing against Intel is not fun. AMD has had only one profitable year since 1995 and sometimes seems to be pressing forward mostly to prove a point. (Think of Rocky III with Mr. T.) IBM doesn't have a personal stake in the issue. It also lost money on PC-compatible chips in a deal with Cyrix in the late 1990s.

Instead, IBM will make money from Opteron by licensing technology to AMD and maybe serving as its manufacturing arm.

Illuminata analyst Jonathan Eunice best summed up mergers with his take on the Sperry-Burroughs deal. The company's new motto was "the power of two," a reference to exponential growth. Eunice called it "the power of two and five-eighths."

5. Sun Microsystems is a software company.

The hardware-as-commodity theory goes as follows: People buy Sun servers today because of the OS and applications. At the same time, hardware is rapidly becoming a commodity. Ergo, Sun can succeed by selling Solaris and middleware to run on Intel-based servers. The same theory is applied to storage companies such as EMC.

Technically, this actually is what Sun is trying to accomplish. However, the end result won't be as clean as expected. Software has to run on something, and if it is not well-integrated into the hardware, it can run poorly. These companies, therefore, will have to keep developing hardware or find a trusted friend to split large projects with.

Few companies have ever successfully changed personalities, and then only because of unusual circumstances. Pixar used to make workstations but was propelled forward by hit movies. Intergraph found a new life after workstations, but only because its patents allowed it to sue Intel, Texas Instruments, Gateway, HP and others.

6. The PC will disappear.

This is the Area 51 theory of the hardware world--it never goes away. IBM was supposed to get out of PCs. Then HP. Internet appliances were supposed to kill them, and soon game consoles and set-top boxes will. Around 161 million will get shipped this year.

In a sense, PCs hold a similar position in the world to cars.

Why don't they fade out? PCs continually get cheaper and the design, despite notable flaws, is functional. IBM and HP may not make much money off them, but if they got out of the market, Dell would swoop in to provide their corporate customers desktops. Every day, there would be a Dell sales rep handing out coffee cups.

In a sense, PCs hold a similar position in the world to cars. Public transportation would be more fuel efficient, but it's slow, and my personal jet pack is on backorder.

7. Services--that's where the money is at.

Everyone is getting into services. But although the market is growing, there is a major problem with services: It's not as profitable as everything else. In the most recent quarter, IBM reported a gross margin of 25.1 percent. That's lower than the margin in hardware, 25.2 percent; software, 85.2 percent; financing, 57.7 percent; and other, 53.4 percent.

Two problems. One, people hate paying for services. With products, you at least point to where the money went. Consequently, service fees get bargained down in contract negotiations. Two, production costs don't go down as fast. Once an application is complete, most of the hard costs are behind you. In services, you're stuck hiring men and women to run around in blue suits to provide company presentations.

8. Web sites tailor their content to suit advertisers.

Every day, I get probably five to six letters accusing me of being a shill for some large advertiser. But believe me, I wouldn't be living in abject squalor if that were the case.

9. Marketers will be able to target consumers via their Web-surfing habits.

Someday this may occur, but the algorithms definitely need some help. That or there is an inordinate number of people interested in mortgage refinancing, alligator jerky, and the secrets of brain transplantation.