Microsoft: let customers decide future

Dateline: April 12, 1998

THIS may not seem AI-related, but in the longer term it is. I've been wanting to say something about the Microsoft business for a long time, and I can't hold it in any longer. It's important. It's also interesting, but doubtless purely coincidental, that twice in the course of preparing this document, Netscape Composer has crashed (I run Windows NT, so my machine stays running.)

The Cincinnati Post of March 27, 1998 reported that Jeff Raikes, Microsoft group vice president of sales and marketing, was in Ohio dutifully touting the party line that customers, and not attorneys for a "misguided" Department of Justice, should decide the future of Microsoft Corp. and its products. There's a bit of flim-flam in that statement.

In the first place, Microsoft customers decide nothing. Did you ask for Windows 98? I thought not. The company calls the shots. Remember Terminal, the awful communication program Microsoft bundled with Windows 3.0? Nobody I knew in the BBS or online services world used it, since there were better communication programs, like Qmodem and many others, readily available. If customers wanted a decent communication program, it was no use talking to Microsoft. Some tried, but were more or less ignored by the company. Communications was a priority for many customers, but not for Microsoft.

Consider also Netscape. The only reason we have a decent browser from Microsoft is not because customers were asking for one, but because Netscape was a threat to Microsoft's ownership of your desktop. That's what Microsoft means when it says "customers decide." If it looks as though you may just take your business out of Windows, then watch the fur fly in Redmond.

In the second place, the DoJ has no brief to decide on Microsoft's future, only to see that it does not abuse its monopoly power. It has the same responsibility for any monopoly, not just Microsoft.

Part of the company’s strategy, Raikes is reported to have said, is to put as many features as it can on its operating systems to give customers the broadest possible range of features. This is especially important, he said, for small business customers who gain competitive leverage against big businesses because computer technology and the Internet "level the playing field."

More smoke. In the first place, it doesn't matter to the small business whether its "competitive leverage" is built into the OS or is attached as an application. In the second place, the issue at hand has nothing to do with big vs. small business anyway.

Microsoft claims that the government’s interference is stifling product innovation in an industry that has helped fuel vibrant economic growth and in which the U.S. now leads the world. Raikes downplayed the results of a recent study by the University of Cincinnati College of Business indicating that the biggest threat to product innovation in the computer industry isn’t market competition, but how successful a company is. If Microsoft becomes too successful, the study suggests, it could cut back on its level of innovation.

It was only to be expected that Raikes would disagree. Any of us would, in his position, I suppose. He said the study did not reflect reality in today’s computer industry where a guiding principle has to be "innovate or die."

Microsoft is certainly very close to death, though not its own. It has all but killed Lotus, Borland, Corel, and many other innovative software makers, by bundling competing word processors, spreadsheets, databases, etc. etc. in with Windows. And now, Microsoft is after the AI market—at least, that part of the AI market that could threaten Microsoft's stranglehold on the desktop.

The U.S. senators who challenged Microsoft Chairman Bill Gates in recent Senate hearings allowed him to rub their noses in it without even realizing it. Gates pointed out, as if in evidence of Microsoft's innovative skills and proclivities, that Microsoft was working to integrate automatic speech recognition into the Windows interface.

Whoa!

Microsoft has been a slouch and a laggard in ASR, just as it was in word processors, spreadsheets, databases, local area networking, and Internet browsing software until others had done all the hard and smart work of innovation and market testing. Microsoft just comes along and, by hook or by crook, "acquires" new capabilities after it is clear they are winners. It has done this in the case of ASR by purchasing the Belgian ASR firm Lernout & Hauspie.

Microsoft's definition of "working" to be "innovative" seems to be simply to buy the innovations from others. It is deeply baffling that Microsoft employs thousands of talented programmers and research scientists, yet the most innovative fruits of their efforts are to be found in the wimpy and thankfully short-lived Bob, Microsoft's misbegotten idea of an intelligent agent.

It has just done it again, with the purchase of Firefly, an AI company whose software provides privacy and sophisticated search filtering capabilities.

Senators, why don't you ask Mr. Gates: When you're through incorporating ASR into Windows, what will happen to Dragon Systems and its fine and innovative ASR programs? What will happen to IBM's equally impressive ViaVoice product? What about Kurzweil's products? What about Fonix? Etc., etc. What will happen to all the other privacy and filtering software innovators competing with Firefly and keeping one another on their toes introducing ever-better versions of their software?

I'll tell you what will happen: they will go the way of Lotus 123, Quattro Pro, WordPerfect, and Paradox: if not quite dead, then almost. That is what happens. But don't expect any Microsoft executive, ever, to admit it.

This would not matter (except to the founders, employees, investors, and satisfied customers in/of those unfortunate companies) if we could be sure that Microsoft would produce a better ASR or filtering or privacy product than its competitors. In the first place, that's not likely, on past performance; and in the second place, even if it did matter, once Microsoft has cornered the market for the product it has but one business imperative: to maximize profits, which will mean to cut R&D and stop delivering really significant revisions.

According to the New York Times of April 10, 1998, Microsoft said it was interested mainly in Firefly's privacy technology rather than its filtering technology, and said its plans for the product were "a little squishy now." Squishy? It is not encouraging to think that Microsoft would spend millions on something it doesn't know what to do with, and that we may have seen the last flicker of innovation from this once bright spark in the fading light of a cold, dark Microsoft night descending upon us all.

In a very small way one might sympathize with Microsoft. An ASR interface allied to intelligent agents with Firefly-like capabilities on your desktop could totally eliminate the need for Windows. Bill Gates sees the threat, alright, and he's doing what most businesspersons would feel constrained to do. The problem is, we're the ones who will be hurt, eventually.

What does one do about it? Somehow Microsoft's monopoly must be ended, but how? There are existing antitrust and other laws, such as the DoJ and several state attorneys-general are currently pursuing, but Microsoft has enough money to keep the legal system tied up till Doomsday. There's been talk of an AT&T-style breakup, in which Microsoft is forced to renounce parts of its business, retaining essentially the operating system part. The problem with that is that Microsoft cannot even be trusted to do a decent job with the operating system. I had a windows-type operating system on my 8086 PC years before MS Windows 1.0—ugh!—appeared. It was called the GEM Desktop, and it was beautiful. It was stable, and it worked. Where are the makers of GEM now? You guessed it.

Most of the world is stuck with Windows 95, an unstable, crash-prone platform. Unix and its derivatives have provided much more stable environments for years. WindowsNT (which I use) is more stable than 95 (though still not as scalable as Unix, in the server version, for large networks), and there's no excuse for not migrating users directly from 95 to NT, instead of via another money-guzzling switch to Windows 98, supposedly to be released this June. But this is what you get, folks, when someone has a monopoly. What is really needed is for the Windows OS to be put in the public domain, as Netscape did recently with its browser.

The alternative to that is for a smart programmer to produce a non-infringing clone of Windows. Such a program already exists in the Unix world. It's called Linux, and it's very good, very popular, and very free. I don't know to what extent (if at all) it can emulate Windows (to let me still run my familiar applications on it) but I am considering switching over to it anyway. There are plenty of good applications written for the Unix/Linux world, it has become much easier to use than before, our Pentium II PCs now have the horsepower that only expensive Unix workstations used to have, and finally, Linux may be the last safe haven for good programmers wanting to produce something of real value without being squished by Microsoft or anything else except a genuinely superior program.

The message for Microsoft in all of this is that it is not going to be business as usual. They have to change. Because ultimately they are right (though for the wrong reasons): the customer will decide their future, once the customer has discovered what the customer stands to lose if Microsoft continues with business as usual.
 
 

  Until next week, 

NEXT WEEK: Possibly, a guest contributor's commentary on my Neuroshell review from three weeks ago. If not, more IMP and EEEK.

Help Wanted: Got questions or comments on this article or on any other AI-related subject under the sun? Post it in the AIBB!

Previous Features