A New Way to Light
While new LEDs are twice as efficient as incandescent bulbs, they don’t come close to matching fluorescent lights (“LEDs vs. the Lightbulb,” TR May 2003). LEDs have a potential light output of 30 lumens per watt, as compared to 10 to 15 for incandescents; but fluorescent tubes that fit into regular light sockets produce 69 lumens per watt. Although it is true that fluorescents will not challenge LEDs for traffic lights or for automobile taillights because of the harsh environments in which they operate, LEDs have a long way to go before becoming common household appliances.
Many of the current efforts to produce white light use phosphors to convert blue or ultraviolet light into longer wavelengths. But how well controlled is the blue itself? The day will come when municipalities consider this technology for streetlights, and it would be criminal to emit as much blue as we do now with our current technology. Besides the unpleasant glare, we would continue to condemn the night sky to invisibility.
More Comfortable Surveillance
The best way to make observations is not by recording enormous amounts of data and then processing it all the time (“Surveillance Nation-Part Two,” TR May 2003). Instead, we should use “smart” surveillance systems that constantly observe, but go into alert mode and record data only when there is a noticeable change from the common pattern of behavior. In this case, there would be no need to record useless amounts of data, and everyone would feel comfortable-unless they were doing something wrong.
North York, Ontario
Making Peer-to-Peer Pay
The record companies just don’t get it (“Curbing Peer-to-Peer Piracy,” TR May 2003). Peer-to-peer can be excruciatingly slow, sound quality can be inconsistent due to poor encoding, the ID3s (data for identifying songs) are often wrong, and file names are not consistent. If I were in the record companies’ shoes, I’d set up powerful server farms on large pipes, collect reasonable monthly fees, and send out high-quality, unrestricted MP3s of every song in my entire catalogue. Give the artists a fair share of the proceeds and watch them rake in the money.
As an inventor with a patent, I read Evan I. Schwartz’s article (“Patents Go Global,” TR May 2003) with interest. He discusses the issue of “first to file” versus “first to invent” as if current procedure can be easily changed. But “first to invent” is a Constitutional right, which will require an amendment to overturn. The definition of inventor includes the idea of novelty, meaning the invention was not previously known. Clearly, if an invention is previously known, the person applying for the patent is not the inventor but has merely rediscovered something already known by someone else.
In Defense of Regulators
Michael Schrage insinuates that regulators, activists, and litigators are Luddites seeking to disrupt the introduction of new technologies without regard to their value (“Global Warning,” TR May 2003). He also proposes that the problems of companies such as Monsanto stem from “consistently lobbing clumsy responses to public criticism.” If only that were true. There are deeper reasons activists and litigators are so active. Independent science conducted by well-funded government regulators is the only ethical way to quiet activists who speak out against new technologies and processes. Such science would also go a long way toward preempting litigation. As for regulators, they are necessary in a world containing growing numbers of people and powerful technologies. The present system is suspect due to the revolving door between agencies and the industries they regulate.
College of Engineering
University of Wisconsin-Madison
Corrections: In the Patent Scorecard (TR May 2003), the company name Magna International was misspelled. Also, Avaya is an independent company that is not owned by Lucent Technologies.
“LEDs vs. the Lightbulb” (TR May 2003) incorrectly implies that a typical incandescent bulb lasts about 8,000 hours. That is roughly the life of a fluorescent bulb; incandescents shine for about 1,000 hours.
Become an MIT Technology Review Insider for in-depth analysis and unparalleled perspective.Subscribe today