Because he's smart, stubborn and very, very persistent, Gilbert Hyatt may avoid ending up as the Elisha Gray of microprocessors.

Gray was the unfortunate soul who submitted his patent application for the telephone a few hours after Alexander Graham Bell. Bell was awarded the patent and won the fame, glory, fortune and sliver of immortality that only a handful of inventors ever achieves. Gray is a footnote.

After 20 years of haggling with the U.S. Patent Office, Hyatt was awarded a broad and comprehensive patent for the invention of "a single-chip integrated circuit computer architecture" -- in other words, the microprocessor. You find microprocessors -- the chip with "brains" -- in everything from VCRs to personal computers to cars. In a way, Hyatt has the patent on the wheel of the information age.

Of course, Hyatt stands to make millions annually if he is able to enforce even a minuscule licensing arrangement with the Intels, Motorolas and Hitachis of the world. More power to him. Every underdog should have his day. The image of the lone inventor toiling in obscurity and beating the Big Boys (like, say, Intel) to the prize is as American as apple pie, Norman Rockwell and MTV.

But what does it say about the way we award patents? Is Gilbert Hyatt's new patent a symbol of what's right with the system? Or does the delay, breadth and surprise quality of the patent reveal some flaws?

"You would like the system to react better than it did in this case," says Gary Hecker, a patent lawyer with the Los Angeles firm of Hecker & Harriman. "There does seem to be something a bit off in finding something out 21 years after being exposed to the greatest possible extent."

Ironically, Hyatt never actually built -- or, in patent parlance, "reduced to practice" -- his microprocessor. You can be awarded a patent for simply describing -- clearly and comprehensively -- how your invention would be built. In fact, Hecker and other patent attorneys point out, many patents filed by companies are never actually turned into products. However, the information embedded in the application becomes public knowledge. That is one way that the patent system fulfills its constitutional mandate to "promote progress of the useful arts."

Even though they wouldn't know a microprocessor from a microwave, the founding fathers were astonishingly insightful to actually make intellectual property protections like patents and copyrights a part of the Constitution. However, the technologies have changed along with the times. It's always difficult to find the appropriate balance between the individual rights of the inventor and the overarching benefits to society.

Precisely because intellectual property is becoming a more and more valuable part of our economy, we need to become more sensitive to those trade-offs. The Hyatt patent provides a window into changes that we might want to consider making in the system.

Certainly, the most shocking aspect of the patent was its surprise. Literally no one was expecting it. There's a reason for this. In Europe and Japan, patent applications are made public 18 months after they are filed. In the United States, they are sealed until the patent is actually awarded. Clearly, this offers American inventors greater secrecy (although many U.S. companies apply for patents overseas and thus disclose).

On average, it takes 18 to 24 months for the Patent Office to process an application. When an application takes as long as Hyatt's, there is no way for anyone in the field to know what impact it might have. Indeed, if even an abstract of Hyatt's application had been released 24 months after filing 21 years ago, semiconductor companies would have had an inkling of what was going on. They could have rushed their own applications or sought to cut a deal with the inventor.

Even though the Hyatt filing is atypical because of its delay, Congress would be smart to pass legislation requiring abstracts of applications to be released within two years of filing an application. It would be a cheap way to let people know what's going on and it would virtually eliminate the heart-attack qualities of the Hyatt patent.

"If we made the patent applications public knowledge," says D.C. patent attorney Peter Trzyna, "that takes the surprise element out of it."

But why did this patent take more than two decades, anyway? Well, if the patent examiner doesn't like the way something is expressed and rejects portions of the application, the inventor can file a continuation. As Hecker points out, Hyatt filed no fewer than seven continuations. That's a lot of bites at the apple.

"He knows how to use the system," says Jerry Goldberg, director of the Patent Office's Computer Group, which approved Hyatt's patent. "He perseveres. Most of Mr. Hyatt's case files are very thick. Very thick."

Patent attorneys love continuations (more money). The Patent Office is less enthusiastic. Inventors should be limited to three continuations. That should help assure that their applications are as crisp as possible and it might discourage some of the broader claims they make. Again, the goal is to strike an effective balance between encouraging and protecting the inventor and assuring that society can reap the full benefits of a new device.

One problem is that the Patent Office is being asked to do more and more with less and less in an era of explosive technological inventiveness. Goldberg's Computer Group has 154 patent examiners to handle a growing backlog of 13,000-plus applications. What's more, the attrition rate of patent examiners is horrendous.

"Last year, we hired 18 people and lost 27," says Goldberg. "We hired 54 examiners this year and we've already lost 20."

A good patent examiner with several years of experience can literally double his or her salary by hopping to a law firm that specializes in intellectual property. Congress should allow longtime patent examiners -- people who have served more than seven years -- to be more fairly compensated. What's more, the Patent Office has trouble recruiting people to work in Washington. The Patent Office should set up a satellite office in California -- where there is a wealth of biotechnology and computer expertise -- and use that as a lure to broaden its appeal. Use video and telecommunications technologies that will let the West Coast examiners perform patent searches remotely.

How to pay for this? Simple. Charge the intellectual property lawyers and companies that file dozens of patents a year a special users fee. That can help pay for the salaries and the information-retrieval technologies the Patent Office so desperately needs. Patent lawyers like Hecker and Trzyna generally have praise for the office -- but they want to see the bureaucracy's own technology catch up with the state of the art.

There will always be difficult and unusual patents like Hyatt's. We should design a patent system that keeps them as exceptions. We literally can't afford them becoming the rule.

Michael Schrage is a columnist for the Los Angeles Times.