It’s time for my predictions of what will happen in high-tech and high-tech business during 2003. But first, as the last honest prognosticator (maybe the only honest one) I have to look at the predictions I made a year ago to see how they came out. If you want to read that old column, you can find it under the Links of the Week on this page.

I said that Microsoft would continue to expand .NET (correct), that XML would be the tool for that expansion (correct) and that KnowNow would be the big XML success story (not yet anyway — wrong). I said there would be a slow resurgence of venture capital (correct) and that would lead to a rise in the number of Initial Public Offerings (well, there was an increase, but it wasn't enough to count — wrong). I said that security technology would be a big deal (correct), but that rich media wouldn't be (correct again). I said that 2002 would be a pivotal year for broadband, whatever "pivotal" means (I don't think it was — wrong). I said that Microsoft would do lots of deals as it attempted to diversify a bit (correct) and that 2002 would be a good year for Cisco Systems (you may not think so, but the stock is up 30 percent from its rock bottom, market share has increased and gross profit margins are the highest ever — correct). That is seven correct and three incorrect predictions, maintaining my historic average of 70 percent right. Whew!

Now back to the future — my predictions for 2003. These predictions may ramble a bit, primarily because I am lately having trouble keeping focused. Blame the lithium.

1. HP/Compaq will continue its long slide to oblivion. It is amazing to read lately all the news stories that say how well Carly Fiorina has done at integrating the two companies when the merger is, in fact, a train wreck. Sales and market share are both down. Everything Walter Hewlett said would happen is happening. The key moment was a boardroom showdown in October between Fiorina and former Compaq CEO Michael Capellas, then the new president of HP. Capellas presented a plan to become more like Dell, building computers to order. Fiorina said it was best to stick with resellers. The board sided with Fiorina and Capellas left the company two weeks later. Smart guy. Given that half the PC retailers in America are under Chapter XI bankruptcy as that industry segment dies, Capellas was right, Fiorina was wrong, and HP is toast. This is a sad end for a company with such a fine engineering and management tradition.

2. Dell gets bigger and bigger. While this is a no-brainer, it introduces another concept that could probably be a prediction in itself, but my list is going to be too long anyway, so we'll just bury it here. That concept is the apparent inevitability of monopolies. Not that Dell will be a true monopoly, but by the end of the year, it will be by far the largest company that solely makes PCs. We seem to have a lot of monopolies these days. Microsoft is one according to the courts. Intel used to be one (more on that later). Sun is certainly the dominant workstation vendor, if that category survives, that is. Cisco has an effective monopoly in networking. We see it, too, in the resurgence of the Regional Bell Operating Companies as they flex their local muscles compared to the pitiful long distance carriers, who have true competition.

3. At the same time, we'll see Microsoft's leadership further influenced by Linux. Remember Steve Ballmer said Microsoft couldn't compete with Linux on price, so they'd have to fight on quality. While you are laughing about that one, I'll further explain that this means Microsoft will have to improve its quality and that may well happen. And if it does, we can thank Linus Torvalds for doing something nobody else, including Bill Gates, could do in the preceding 25 years.

4. If Linux is giving Microsoft fits, it is doing far worse to Sun Microsystems, which I predict will have a very bad 2003. As just one example, Sun is in danger of losing the semiconductor design computer workstation market to Linux. Early this year, Cadence Design will be the last of the major vendors to port their software to Linux. In the server market, too, Linux is making real inroads at the expense of Sun, especially in 2003 as the 64-bit Linux boxes begin to appear. Why buy a $100,000 Sun server when a $10,000 Linux cluster is comparable in every way? And don't expect too much from Sun's own Linux boxes, which will be deliberately hobbled so they don't make problems for SPARC.

5. Meanwhile, China, which will eventually be the largest computer market on earth, will standardize on MIPS processors and Linux, much to the dismay of both Sun and Intel. This bodes well, by the way, for AMD with its new MIPS-based Systems-on-Chip that will be the major component in many of those el cheapo Chinese computers.

6. And as I said last week, Microsoft will force Intel to Adopt AMD's Opteron 64-bit extensions. There is a small possibility Microsoft will actually buy AMD, which would really shake up the semiconductor market, but I think this is unlikely.

7. In the meantime, Apple will announce a line of computers using IBM's incredible Power 4 processor, but of course, they won't actually be delivered until 2004.

8. The V.92 modem standard still won't work, but nobody will care.

9. Microsoft's Palladium security initiative will become even less popular as people realize that Palladium is too intrusive and it doesn't really work. A year isn't enough time for Palladium to die, but it eventually will.

10.Still, security will be a bigger thing than ever, though you'll see a subtle shift from people being worried about viruses to them being furious about spam.There is a key issue here, and it is that spam makes money while viruses do not, so spam is going to get only bigger while more and more people are successfully dealing with viruses.Unsolicited e-mail will be a bigger problem than ever.

11. I wish it wasn't so, but by the end of 2003, I am sure we'll see at least four new laws giving corporations the right to invade our privacy, with most of those laws passed in the name of "patriotism."

12.Hollywood will come up with another new copy protection scheme for music and it will be defeated within two months. Even more significant is the fact that 2003 will see a whole new generation of peer-to-peer file sharing software.Remember, Napster required a central server and Gnutella replaces that with a zillion individual PCs, but this next generation will probably be some form of swarm computing — a distributed network file system that will chop up the music into a hundred pieces each stored on a different participating PC.So when the RIAA confiscates your computer, they won't be able to find anything stored on it resembling music.

13. 802.11a loses out to 802.11g.

14. I'm sorry to be mysterious about this one, but I seriously doubt that there will be a significant act of high-tech terrorism in 2003, yet I think there will emerge an entirely new way to use the Internet to effect social change. Just as Chinese fax machines made inevitable the events at Tiananmen Square, so the Internet will bring the world's attention in an entirely new way to social issues — much to the dismay of the status quo.

15. And finally, with the continued (and to me totally inexplicable) rise of web logs, someone — maybe Google — will come up with an effective blog search engine to read all that junk for us and extract what we really care about.