brolston + innovation_&_science   2

The Supreme Court Should Invalidate Software Patents
Image via Wikipedia

Last weekend I was thrilled to hear one of my favorite radio programs, This American Life, take up the issue of software patents. Computer programmers have been sounding the alarm about this problem for two decades, and it’s great to see mainstream media outlets finally start to give the issue the kind of attention it deserves. TAL devoted a full hour to the subject, focusing on Intellectual Ventures (which I’ve written about at length) and did an absolutely spectacular job.

This American Life‘s story-telling format makes it great for describing a problem, but it didn’t spend any time discussing potential solutions. So in this post I hope to fill in the gap by describing what I believe to be the best solution and how we ought to get there.

In my view, the solution is straightforward: software shouldn’t be eligible for patent protection. That might sound simplistic, but there are good reasons to think abolition of software patents is the right reform. Software is fundamentally different than other types of inventions. For starters, software is virtually alone in being eligible for both patent and copyright protection. This makes patent protection mostly superfluous. Second, writing software is an individual, expressive activity at least as much as it is an engineering discipline. We don’t expect novelists to hire patent lawyers, and computer programmers shouldn’t have to either. Finally, the “software industry” is radically more diffuse and diverse than the typical patent-eligible industry. Every business with more than a handful of employees has an IT department producing potentially patent-infringing software. No other category of patents has this characteristic.

Unfortunately, as Matt Yglesias points out, the patent reform legislation now working its way through Congress is woefully inadequate. I’d love to think that a wave of negative publicity for software patents would produce better legislation, but that’s not realistic. At this point, software patents simply benefit too many entrenched interests to expect Congress to enact serious reforms.

That means that the best hope for reform lies with the courts. The Supreme Court said three times that mathematical algorithms (a.k.a. “software”) are not eligible for patent protection. Unfortunately, the last of these decisions was three decades ago, and it was muddled enough to allow lower courts to gradually make software patents easier to get.

But in principle, those old Supreme Court decisions are still good law, even if lower courts have gotten in the habit of ignoring them. The Supreme Court just needs to say they really meant it. Indeed, many software patent critics hoped that last year’s Bilski v. Kappos case would be a first step in that direction. The case focused on “business method” patents, which was legalized in the same 1998 decision that decisively legalized software patents. But the Bilski case wound up being a 5-4 nail-biter, with a conservative majority striking down the particular business method patent in the case but refusing to rule out business method patents in general.see photosForbes ImagesClick for full photo gallery: A Visit To Xerox PARC

There is strong circumstantial evidence that Justice Scalia was (uncharacteristically) the swing vote, and that he is deeply ambivalent about business method and software patents. In his majority opinion, Justice Kennedy worried that invalidating business method patents would “create uncertainty as to the patentability of software, advanced diagnostic medicine techniques, and inventions based on linear programming, data compression, and the manipulation of digital signals.” That would be a pretty depressing read for software patent opponents like me except that Scalia pointedly declined to join this part of the majority opinion. With only four votes, that part of the opinion isn’t binding precedent. And this isn’t the first hint that Scalia has doubts about software’s patentability.

Justice Stevens wrote an impassioned dissent calling for business method patents to be invalidated. Stevens retired shortly afterwards (see my tribute), but his dissent was signed by the three other liberals still on the court. We don’t know what Justice Kagan thinks, but it seems likely that she would have sided with her fellow liberals. Four liberals plus Justice Scalia would be a majority.

Of course, invalidating software patents at this point would be intensely controversial, because it would invalidate hundreds of thousands of patents—worth billions of dollars—at a single stroke. Courts always try to avoid upsetting apple carts. But in this case, invalidating those patents would be good policy in addition to good law. The growing value of software patents represents not the production of new wealth but an increasingly lucrative form of rent-seeking. As the number and value of software patents grows, the case for invalidating them gets stronger, not weaker.

But the Supreme Court won’t take such a dramatic step unless there is a broad consensus that patents are detrimental to software innovation. And this is why it’s so valuable to have mainstream programs like This American Life covering the issue. Justice Kennedy was obviously unaware that most computer programmers consider patents an impediment to their work. Only if this fact becomes common knowledge, in the way that everyone knows doctors hate malpractice lawsuits, will we have any hope of the Supreme Court—and specifically Justice Scalia—doing the right thing.
Innovation_&_Science  Law  Op/Ed  Policy  Tech  Bilski_v._Kappos  Business_method_patent  Intellectual_Ventures  Patent  Supreme_Court  This_American_Life  United_States  United_States_Supreme_Court  byline=Timothy_B._Lee  from google
july 2011 by brolston
A Crisis in Reporting?
Image via Wikipedia

Clay Shirky is one of my intellectual heroes (see my 2008 interview with him), so I read his latest post on the future of reporting with interest. He says that as readers have shifted online, newspapers have suffered from an “analog dollars to digital dimes” problem: the amount advertisers are willing to pay for online readers is dramatically less than what they used to pay for print readers. Shirky argues that news organizations are going to have to learn to live with dramatically lower revenue per reader.

He makes some sensible recommendations for how the news industry and the broader society should deal with the situation. But what stuck out for me was Shirky’s uncritical endorsement of the conventional wisdom that we’re in the midst of a grim “crisis in reporting,” which, he says, “isn’t something that might happen in the future. A 30% reduction in newsroom staff, with more to come, means this is the crisis, right now.”

Shirky is sometimes criticized for the rose-colored tint of his spectacles, but here I think he’s giving too much credence to the pessimistic conventional wisdom. It’s clear that newspapers are facing a crisis, and obviously if you’re a newspaper employee or shareholder you should be worried. But whether this is a problem for the broader society is far from clear.

To tell whether the decline of newspapers is just a normal story of disruptive innovation or something the rest of need to worry about, we need to look at outputs, not inputs. It wasn’t a “crisis in telephony” when the switch to automatic dialing allowed AT&T to lay off thousands of phone operators (unless you got laid off). It wasn’t a “crisis in computing” when the PC put minicomputer manufacturers like DEC out of business. By the same token, a 30 percent decline in newsroom staff might just reflect increased journalistic efficiency.

How could this be? A reporter’s job is to collect, organize, and summarize information about important events in the world. The more of the world’s information that comes pre-organized, the easier the reporter’s job will be. And the Internet is a gigantic information-organizing machine. Sites like Google, Wikipedia, and Twitter provide vast amounts of information “pre-digested.” Reporters still have to do some work to get the information they need from these tools and turn them into publishable copy. But more and more of the basic work is done for us.

For example, in my last post, I included some statistics about Microsoft and Google’s patents. Because the patent office has a searchable online patent database, this only took about 20 minutes. I also save a ton of time any time I’m covering legal stories because statutes, court opinions, and other primary documents are usually available as PDF downloads. Similarly, Matt Yglesias regularly writes posts like this where he graphs some kind of data pulled from a government website. This kind of information retrieval is now so easy that we barely even think of it as reporting. But in the pre-Internet world it would often have been a much bigger undertaking.

Another example: I get many of my story ideas from Twitter. Every couple of weeks I do a tweet like this asking what I should be writing about. More often than not, one or more people reply with great suggestions. I also follow a number of activists, academics, and think tankers who do work related to my beat. Their tweets frequently give me story ideas. As a result, I spend less time hunting for story ideas, and more time actually writing them.

The Internet is also reducing duplication of reporting effort. The 20th century newspaper industry had a lot of reporters covering identical beats in different cities. Obviously, each metro area needs its own reporters covering city hall. But a ton of stuff in the newspaper—technology and medicine, national business and politics, movie and book reviews—isn’t tied to any specific metropolitan area. As the Internet eliminates geographic boundaries, there’s no longer a good rationale for having so many people writing redundant content.

Another important way the Internet makes reporters more productive is by reducing bureaucracy. A significant part of a newspaper reporter’s job involves negotiating with her editor about which stories she should write, when they’ll run, and how much space they’ll get. Print reporters sometimes waste time on stories that get spiked, file under-reported stories to meet arbitrary deadlines, or cut out interesting material to save space.

Shirky has called this a “filter, then publish,” process. In contrast, Forbes bloggers like me operate on a “publish, then filter” model. We write whatever we want, the Forbes editors decide which content to promote on the Forbes home page, and we’re paid based on the traffic we receive. This is more efficient not only because we don’t have to waste time negotiating with our editors, but also because there are fewer perverse incentives: we get paid if and only if we write stuff people want to read.

I could go on, but you get the idea. Add all of these factors up and I think it’s entirely plausible that the news industry’s productivity has improved enough to offset that 30 percent fall in newsroom headcounts.

Many of us look back at the 1970s and find it hard to imagine a world with just 3 or 4 national television networks. I suspect that in the 2030s, people will look back at the 1990s with the same kind of astonishment that people were satisfied with the limited amount of news available from a single newspaper. So why do so many people today see the decline of monolithic newspapers as a calamity rather than a sign of progress? Partly it’s cultural inertia, but I suspect it also has something to do with the fact that the companies whose oxes are being gored own some of the nation’s tallest soapboxes.
Business  Economics  Innovation_&_Science  Media_&_Entertainment  Op/Ed  Tech  byline=Timothy_B._Lee  from google
july 2011 by brolston

Copy this bookmark: