Sunday, August 10, 2014

"Patent or perish" and the story of "Gorilla glass"

An article Patent or perish? Jeers greet proposal to tie Australia's academic science funding to patents describes largely negative reactions to a proposal made by Ian Macfarlane for
"rewarding them [block grants] to universities not on the basis of how many papers they've had published, but actually on the basis of how many patents they've had registered." Additionally, MacFarlane noted that the current grant system emphasized a researcher’s productivity in publishing technical papers, “which are great if you're into producing papers, but I'm into producing jobs, and I'm into producing products that are commercialised from the IP [intellectual property] that that scientist or researcher may have developed.

Given the relative ease in getting a patent in Australia, Macfarlane's proposal is not as onerous as it would be in the United States. Separately, the tie to commercialization evokes the Bayh-Dole Act in the United States, which was somewhat controversial in the 1980's but now has a firmly established constituency. And, recall even further back, the tremendous opposition of doctors to Medicare, which later became a gold mine for doctors. Early criticizers can become staunch proponents later.

The idea of MacFarlane that, if the public funds the research, the public should get some return, is not so unreasonable. California's boondoggle in stem cell research would be a case in point.

Buried in the comments of the "Patent or perish" article is a reference to Gorilla glass (once called Chemcor). From the Gorilla glass comment:

Touchscreens' Gorilla glass was Chemcor lithium aluminosilicate, a miracle without applications. 40 years of research diddled surface hardening and toughening. Only a fool would have sustained it. Gorilla Glass III is fat money, and irreplaceable...

...until somebody is cleverer. HR excludes all but mediocrities plus diversity hiring. Korporate Kulture calls discovery "insubordination." Pyrex-brand is no longer Pyrex glass. Managerial performance assessments assassinate the future.

The story of "Gorilla glass" evokes inventions of Xerox which were not so valuable in the hands of Xerox, but quite valuable in the hands of others [sublimely ironic in view of the history of Chester Carlson]. This brings up a point about the patent system, discussed in a thread about "Gorilla glass":

Hypothetically, this is what patents are actually for. By describing your process to the point that anybody could replicate it from the patent alone, the government grants you a monopoly, and once that is up it theoretically never disappears from public knowledge again.

Realistically there are issues of course, but that's the theory.

(This is actually one of my core reasons to be against software patents; even if they otherwise worked as designed, in practice a lot of software patents are ten pages of legalese, half or more dedicated to simply enumerating the presence of a computer in the invention and various combinations of "on a network" and "with a disk drive" etc, and in the end the patent document is usually more of a spec then an implementation. If we really carried through on the logic of software patents, a software patent ought to require a software implementation in clear source code that is proved to implement the desired algorithm within a reference environment, and the source code should itself have no copyright protection. Combine that with a shorter term and I'd at least object much less, though I ultimately do agree in abolition in this domain; trade secrets and copyright are adequate to the task, if not already overkill.)

Somewhat more directly on topic, I bet 1960s and 1970s material science patents probably hold a few more things that are useful today that weren't useful then.


Of course a patent still wouldn't have helped in this case if it has been around for 50 years.


No, that's the exact opposite of my point. By patenting something, you stick in it the database of "public knowledge" so that it could be discovered 50 years later if need be. That means it wouldn't be necessary for Corning to sit on it for 50 years. Sitting on things for 50 years and making the current status of the filer irrelevant (dead, disincorporated, whatever) is one of the purposes of the patent system.

The "limited monopoly" is part of a social contract and understanding the whole contract is important.


Post a Comment

<< Home