3. Gladwell responds to Pinker.
4. Via Felix Salmon, Clay Shirky on ontological classification.
6. Why does the universe look the way it does?
7. Inventing a better patent system.
8. Interview with a Michelin inspector.
Assorted links
Previous post: San Antonio bleg
Next post: Personalized markets in everything, increasing cost edition
I liked link #7 much better than your last Assorted Link to a patent-related article. I thought myself somewhat informed on patent issues, but I was totally surprised by one aspect of patent law highlighted in the article:
In most foreign countries, inventors with trade secrets are protected from infringement suits if they have been using the secret innovation for some time, another firm independently develops a similar innovation and obtains a patent for it. This protection is called “prior user rights.† In 1999, Congress limited prior-user rights to “business method† patents like Netflix queuing; the statute did not cover prior uses of machines or other actual products. Congress should extend protections to prior users of all types of patents.
Patent people, what WTF is going on here? If Party A is practicing an invention before Party B patents it, how can Party B ever patent it in the first place? “Unobvious to one skilled in the arts”? For what definition of “one” and/or “unobvious”?
Shirky doesn’t seem to realize that ontologies are not strict hierarchies. An individual can belong to multiple classes. So the problem he mentions with Yahoo is not really relevant.
The answer to Curt Fischer is that the patent system is designed to push inventors to disclose their knowledge in exchange for limited exclusive rights — that’s the so-called “quid pro quo of the patent system.” If Party A is practicing an invention secretly before Party B independently invents and discloses to the public in exchange for a patent, then Party B is entitled to a royalty from Party A.
But isn’t the idea that the knowledge disclosed is supposed to be novel? I can think of at least “one” skilled in the art for whom Party B’s invention was obvious. I don’t see why Party B should be entitled to Party A’s money when Party B was first to invent.
But if the system in place grants Party Bs those rights, then I strongly favor maintaining the US’s first-to-invent rule instead of the first-to-file rule, as long as Party Bs can demand money from people who invented their product before them. What is the public good in forcing everyone to file as soon as they are physically able?
Gladwell made a royal fool of himself in that “rebuttal”. Someone made up a fake statistic and took Gladwell along for the ride, and now he can’t admit that he was being conned. Instead he retorts to name calling. Pinker, and subsequently Sailor, owned him pretty hard. Can’t be fun for him to be so outdone by Sailor…
Curt,
An invention is considered novel if it has not been publicly disclosed. Secret knowledge does not prevent something from being novel, that would completely defeat the purpose of patents.
If you invented first but decided to keep your invention secret, too bad, now you owe the guy who did patent it money (not that I would particularly object to an exception for that first inventor). Also, there is an advantage to making inventors file as soon as possible, and that is the advantage of the knowledge getting out there to others as soon as possible.
The real “problem” with patents is fundamental. They are not set up appropriately to achieve the goals they are meant to. The main purpose of patents is to encourage costly research and innovation where imitation is inexpensive. However, the 20-year patent term is the same regardless of how much money the research cost or how expensive imitation is.
So, you see software/process patents granted for things that took someone 20 seconds to think up- which they probably would have done regardless of the existence of patents. But the way the patent laws are set up, if something is new and nonobvious, cost/effort does not come into play. So a lot of people want to exclude software from patentability for that reason, but there is no justification for such an exclusion under the law.
The New Yorker article on the Michelin guide translates the title of the book L’Inspecteur Se Met à Table overly literally as “The Inspector Sits Down at the Table†. In fact, “se mettre à table† is a standard idiomatic expression in French meaning to “talk” (in the sense of giving a confession to the police), to admit everything. So there is a wordplay there that is lost in translation.
Here is a good article on drafting QBs:
http://www.advancednflstats.com/2008/04/drafting-qbs-2.html
The author finds that there is actually no difference in terms of performance for QBs after the first two picks. That is, the first QB chosen is much better than the second, and the second is somewhat better than the rest, but after that the teams seem to be drafting QBs randomly.
So in some sense Gladwell is right and in another he isn’t.
Andy,
I suggest you reread that article. The author does not say what you say he says, nor does the data he cites show that.
Gladwell’s rebuttal is also kind of stupid b/c it’s not like there’s an entire profession that publicly publishes football analyses the way there is for econ or neuroscience. It’s totally legit to go to non-peer reviewed pubs when that’s the bulk of what there is. Does anyone really believe that there are as many 8th and 1st round QBs who are successful? Plus, the draft system works against the very top picks by sending them to the worst teams, which could dampen the performance of very high picks.
Just looking at today’s QB leaders. Here are the QBs/rounds: Favre (2), Brees (1), Manning (1), Rodgers (1), Brady (8), Schaub (3), Rivers (1), Rothlisberger (1), McNabb (1), Romo (undrafted), Warner (UD), Orton (3), Flacco(1), Manning (1), Palmer(1). Is that 9 out of 15 that are first round picks?
Malcolm Gladwell is America’s foremost pseudo-intellectual. Tyler, you ought to direct readers to the comments section on Gladwell’s post, where Sailer thoroughly shreds Gladwell’s methodology.
Comments on this entry are closed.