TCS Daily


Glasnost in Armonk

By Michael Rosen - October 19, 2006 12:00 AM

MENLO PARK, CALIF. - Google made big news here earlier this month with its $1.65 billion swallowing of YouTube. The acquisition evoked an earlier epoch when the Dow flirted with 12,000 and startup companies (like Google itself) could float or sell themselves for astronomical sums.

But while this new-old-school purchase made big headlines, another development in the tech world has been traveling under the radar even as it portends potentially significant changes in the way that technology companies deploy their intellectual property.

In late September, IBM - arguably the oldest tech startup still around - took a major foray into the related fields of open-source software and patent reform by announcing that it would make its patent applications available for public scrutiny.

This marked only the latest wrinkle in Big Blue's increasing willingness to "liberate" its patent policies. Back in January 2005, the Armonk, NY-based company dedicated over 500 of its patents in the software field to the public.

While these composed a miniscule fraction of the 10,000 software patents in the company's portfolio, the move nonetheless signaled Big Blue's growing affinity for the (free) Linux platform and for an operating system-style showdown with Microsoft; by making some of its computing IP freely available to software developers, IBM hoped to encourage others to produce killer programs capable of running on Linux and challenging Windows.

The maneuver earned predictable plaudits from open-source types secretly hungry for the kind of recognition that only a major company can bestow.

And bestow IBM most certainly has. Its latest gambit appears to be even more ambitious than its previous surrendering of patents. According to the company's press release, IBM will relinquish another 100 patents in the controversial "business methods" field, an area encompassing Amazon.com's (in)famous "one-click" patent. The company announced its official view that "pure business methods [inventions] without technical merit should not be patentable."

Big Blue also plans to make its applications available for all to review within 18 months of filing them. Any programmer or academic out there who believes that certain similar products or patents predate IBM's claimed inventions will be encouraged to contact the company and share this "prior art."

The company also promises to provide its own "technical experts" who will "spend thousands of hours annually" sifting through other companies' patent applications in an effort to determine whether prior art exists.

The guiding principle behind the policy is that "patent quality is a responsibility of the applicant," according to the company's senior vice president for technology and IP. The new approach was apparently midwived by a two-month-long online wiki forum in which a "worldwide community of 50 experts in the fields of law, academia, economics, government, technology and others...collaborated with IBMers to...establish a blueprint for meaningful change."

On the whole, this new policy is a reflection of the company's belief that "widespread adoption of a more formal code of conduct around patents could ease the burden on legal and government administrative systems." It's an attempt to ensure that the patents that do end up issuing are clear, well-founded, and unimpeachable, a highly laudable goal.

Of course, some have interpreted these moves as yet more evidence that something's rotten in Armonk. IBM has gradually extricated itself from what was once its core business: making computers. In 2005, it stopped selling PCs, instead licensing its popular ThinkPad notebook line to Lenovo, a Chinese company that now manufactures all IBM-branded computers.

Ostensibly, these divestments were aimed at enabling the company to focus on its core competence: developing software. Yet now, the world's leading intellectual property company - IBM's portfolio reportedly exceeds 25,000 in the U.S. and 40,000 worldwide; it received 3,248 issued patents in 2004 alone - appears to be dismantling its own empire, prompting some analysts to wonder: what gives?

I think there are two equally plausible explanations: a cynical one and an altruistic one.

First, let's indulge the cynics, including one patent law expert that I interviewed. For one thing, it's not yet clear exactly which of its patents IBM considers to be illegitimate business method inventions that are unworthy of exclusive protection and which, instead, it intends to continue relying on.

It could also be argued that by presenting its applications to the public (and generally not sooner than 18 months after they're filed), IBM is merely outsourcing its own patent review efforts to the community of programmers. Why pay an internal team of lawyers and technicians when slashdotters galore will gladly review your patents for you?

And cynics might also see a sinister motive in Big Blue's announcement that it will now be monitoring PTO filings for dodgy patent applications filed by others. Is the company actually committed to ensuring the objective validity of all software patents or is it just trolling (perhaps "goating" would be more appropriate) for shaky patents that might inhibit its business?

But on the other hand, from an altruistic perspective, in implementing its new policy, IBM could be exploiting the company's first-mover position to foster positive changes across the industry. As the Hummer of the computing IP world, guzzling Patent Office resources like they're going out of style, Big Blue has opted for fuel efficiency, shrinking itself from the H2 to the H3. In so doing, IBM also appears to hope that its competitors will follow suit, thereby conserving the finite supply of PTO time and money. Even if the policy is to the company's immediate detriment, it's betting that it will influence the system enough to restore balance.

Thus, the company seems to be adopting the approach of Harvard College, which recently announced that it would no longer accept "early" applicants. While at first blush, the decision would appear to disadvantage Harvard relative to its competitors for elite students (such as Stanford and that school in New Haven) which will now get first crack at top high schoolers, the nation's oldest university felt confident enough in its position that it could sacrifice its own well-being for (what it saw as) the good of the system (Harvard officials believe that the early admission process burdens students who require financial aid, since by considering only a single school, they cannot compare aid packages from multiple colleges).

Interestingly, Princeton and other schools have followed Harvard's lead in abolishing early admission while all admissions directors across the country have had to engage in the debate.

So too, IBM seems to expect that that which will benefit the industry will also bolster its bottom line.

In a sense, then, the two explanations for IBM's actions dovetail: the company is undoubtedly trying to improve its competitive position while also attempting to improve the system for all players. Big Blue is quite possibly trying to do well by doing good.

Whether or not the plan will work remains to be seen. But this could prove to be one of 2006's most important developments in the tech and IP worlds, the Google grab notwithstanding.

Michael M. Rosen, TCS Daily's Intellectual Property columnist, is an attorney in San Diego.

Categories:

11 Comments

The Most Interesting Idea Here. . .
. . .is that IBM might be gearing up to make a challenge to Microsoft's operating system monopoly. And IBM actually has the size and pull to make a threat of it.

Could be very interesting times in the computer world.

Patents a problem for IBM
I think IBM feels a bit inhibited by all the patents in the market.

It is a well known fact that some companies seek patents for protection against lawsuits from competitors with patents rather than getting exclusivity on an invention.

IBM could be one of those companies. If they could get other companies to follow their example, IBM will be more free to develop grat software.

Everybody (almost) can get a great idea, but only a few can make it work reliably - and IBM can do that.

Go Big Blue
It would be nice to see real competition in this area.

Cynicism...
IBM appears to be merely making its own good publicity without any risk. Here's why:

First, most patent applications are published 18 months after filing. Thus is required, by treaty, if the applicant is going to seek foreign patent rights, which I assume that IBM typically does. Thus, making an application available for public review after 18 months is no concession to the public.

Second, offering the public to provide prior art that might invalidate an application could be a great benefit to IBM. Such prior art could be used to invalidate a patent if IBM should try to enforce it, so IBM would naturally desire the best prior art during prosecution to ensure a valid patent.

Now, it is true that some patent trolls don't care if the patent is valid or not, as they rely on the threat of litigation to force royalty payments. But IBM could not operate that way, due to the poor publicity it woulld get. Furthermore, many of IBMs competitors are big companies that can defend themselves, and are likely to do so. Thus, IBM wants strong, valid patents, and would gladly seek the public's help.

Regarding Business Method patents, the IBM position is mostly a default anyway. The USPTO, in practice, is typically requiring some kind of technical innovation anyway. Business methods that are manually implemented are very difficult to get today. Thus, yet again, IBM is relying on default practice to get good publicity.

As to donating some (small number) of its patents to the public, I have no personal knowledge of the technology covered by those patents, but I suspect that they are technologies that IBM has chosen to forgoe, but would like to see in the public domain so that it can take advantage of any resulting products (and the low prices that would result from any competition that might occur). If IBM had thought that the patents had any commercial value, it would have been a travesty to their shareholders to give them away, and a shareholder revolt would be expected.

Finally, regarding open source, any smart company would like to see an alternative to MS Windows. Competition is always good. And considering that MS built its empire based at least in part on IBM financing and technical expertise, and IBMs poor business judgment (and perhaps a poor assessment of the shrewdness of Microsoft--i.e., Bill Gates), this issue is a no-brainer.

In essence, I don't see any real concessions here. Instead, I see a company that is shrewdly manipulating the press to make it appear that they are doing something in the public good, when the reality is that they are doing what is either best for them, or at least already the rule, and thus not at all detrimental. These actions are hardly a dismantling of any empire.

A pretty smart move on IBMs part, if you ask me, considering how gullible the press is.

-Bob

So what?
Leclerc thinks that IBM feels inhibited by all the patents in the market. But IBM is the patent leader. How do you think their competitors think?

Furthermore, Lecelrc says that it "is a well known fact that some companies seek patents for protection against lawsuits from competitors with patents rather than getting exclusivity on an invention". So what? The patent only has value if it discloses a useful invention. Yes, companies do patent improvements in order to force patent holders into licensing arrangements. That is the free market at work, and it reduced the negative impacts of monopolies. One can only get a patent on a NEW idea, so a competitor can't run out and get a patent for an IBM product. It must "invent" something new. There is nothing negative about defensive patents, as long as they are valid (i.e., not junk).

Leclerc also says that if IBM "could get other companies to follow their example, IBM will be more free to develop grat [sic] software." Why? IBM has far more patents than any competitor. Of COURSE they would like their competitor to have fewer. But IBM uses the patent system to its advantage. So should everybody else. Anyway, history likely shows that IBM has probably developed less "great" software than smaller companies have. So IBM is no saviour in this respect.

Finally, Leclerc says that "[e]verybody (almost) can get a great idea, but only a few can make it work reliably - and IBM can do that." But IBM has had quite a few flops, and others have shown much greater ingenuity (note that IBM farmed out much of its PC operating system work to Microsoft). Yes, it is difficult to get a great idea to make money. But IBM is not necessarily better equipped to do so, if history is any guide.

-Bob

That's why!
"Leclerc thinks that IBM feels inhibited by all the patents in the market. But IBM is the patent leader. How do you think their competitors think?"

It soes not matter how many patents you have. If you have a million patents, I can still inhibit your business with one single patent. I can even stop you from using some of the patents you have! Leader or not, it does not matter.

"Furthermore, Lecelrc says that it "is a well known fact that some companies seek patents for protection against lawsuits from competitors with patents rather than getting exclusivity on an invention". So what? The patent only has value if it discloses a useful invention."

That is not intirely correct. Software patent is such a jungle that you don't know that you have violated somebody elses patent until the sue you for it. That is a huge liabity. So a company with a patent portfolio makes an agreement with other companies that they don't sue each other for violations by them. Some only maintain a patent portfolio for that sole purpose. Hence it is the patent system that makes patents necessary.

"One can only get a patent on a NEW idea, so a competitor can't run out and get a patent for an IBM product."

True, but IBM can't use an idea IBM has come up with by themselves if somebody else has patented it. That is a serious problem for IBM.

Again, So What?
Leclerc says: "It soes not matter how many patents you have. If you have a million patents, I can still inhibit your business with one single patent. I can even stop you from using some of the patents you have! Leader or not, it does not matter."

If I have a million patents, I have a million times more capability to inhibit a business. IBM obviously considers patents valuable (they cost them a LOT of money). You think that IBM isn't using its patents to inhibit others? Then you live in a dream world.

Leclerc also says: "Software patent is such a jungle that you don't know that you have violated somebody elses patent until the sue you for it. That is a huge liabity".

First, software is no different than any other category of invention. Any inventor is at risk that his invention has already been patented. Furthermore, it is not necessarily true that you "don't know" if you infringe a patent. A software engineer can perform a patent search and claim analysis just like any other inventor, or pay a patent attorney to do it for him.

Leclerc says that "a company with a patent portfolio makes an agreement with other companies that they don't sue each other for violations by them. Some only maintain a patent portfolio for that sole purpose" and I again say, so what? How is this "wrong" (as long as its done without infringing fair trade or competition laws). He goes on to say that "it is the patent system that makes patents necessary" to which I say "what?" Of COURSE patents would not exist without a patent system. This is an obvious truth. It is also not relevant.

Finally Leclerc states that "IBM can't use an idea IBM has come up with by themselves if somebody else has patented it. That is a serious problem for IBM", to which I again say, "so what?" Are you trying to say that there is something wrong with IBM having to follow the same rules as everybody else? Well then, too damn bad for them!

-Bob


Whats why, again!
"If I have a million patents, I have a million times more capability to inhibit a business."

This is not true. Key patent can prohibit the use of countless patents. A typical piece of software uses hundreds if not thousands of patents. If I own just one of these patents, I can inhit the entire software completely.

"You think that IBM isn't using its patents to inhibit others? "

Of course it does. That is the name of the game, but it does not justify the existence of the game. The problem is the purpose of having patents is surviving the patent system. In that situation everybody will be better off without any patent system at all or with a more refined system..

"Furthermore, it is not necessarily true that you "don't know" if you infringe a patent. A software engineer can perform a patent search and claim analysis just like any other inventor, or pay a patent attorney to do it for him."

The cost of doing that is in it self prohibitive. Software patents are very difficult to understand even for programmer - and that is a big difference. And when the number of patents in a single piece of software is in the hundreds, it is impossible do a proper search.

" Finally Leclerc states that "IBM can't use an idea IBM has come up with by themselves if somebody else has patented it. That is a serious problem for IBM", to which I again say, "so what?"

SO WHAT!?! You don't find it a problem that IBM cannot use their own idea they came up with themselves because somebody else did it before? This is a huge contradiction in IP law because the basis of IP is that you can own your ideas. Well in some cases you cannot. And that often happens I software.

Ahah!
Now we get to the truth, Leclerc is against the patent system in general, claiming that the only purpose of the patent system is its own survival.

However, what Leclerc misses, is that there is a clear link between a strong patent system and a strong economy with much innovation. Those countries with weak systems are not associated with innovation.

Furthermore, the basis of IP law is NOT that one can have their own ideas. Instead, it is to reward those who innovate and take risks with a limited monopoly in exchange for full disclosure. The world benefits from the disclosure, but can't sell the invention for about 20 years.

Regardless, Leclerc exagerates the risks to programmers. I know, as I help to advise them in such matters.

-Bob

No Subject
The US patent system only got going *after* US companies had intellectual property to lose. Before the end of the nineteenth century, the US attitude to IP was similar to China or Russia's now.

Or to put it another way, it may be that strong economies develop strong systems of IP protection, not the other way round.

Ahah, so?
I'm not against the patent system as such, but I am not convinced about the unconditional rightness of the system.

I agree there seems to be a link between strong patent systems and strong economies. However the link could also be between strong economies with innovation and non-corrupt states.

I think it is a serious problem if the system is not about assigning property rights to your own ideas. Your "world benefit" argument is utilitarian when it is worst, and I do not like it.

All in all I do not like monopolies, especially those founded in legislation. Some monopolies can be tolerated for the lack of better options. I do not think you are making a great case for one of those.

TCS Daily Archives