Aurilink Blows Smoke

In their latest press release, Aurilink Inc. claims that their new amplifier makes customized digital hearing aids obsolete:

Aurilink, Inc. has developed a “ready-to-wear” sound amplifier for individuals who want and need occasional hearing assistance due to mild hearing loss. Such assistance was previously only available with expensive custom hearing aids, but preliminary studies have shown that the Aurilink Sound Magnifier is equal or superior to the most advanced DSP (digital signal processing) hearing instruments currently available.

As part of the team that develops and produces some of the world’s most advanced DSP platforms for hearing instruments, I’m skeptical, to say the least.

Hearing aids are very similar to glasses. Just as people have different kinds of vision problems — near-sighted, far-sighted, and astigmatism — and to varying degrees, they also experience different kinds and varying degrees of hearing loss. A custom hearing aid doesn’t just make everything louder, it amplifies the specific frequencies where the wearer has suffered loss. That’s what makes them customized. Beyond that, high-end digital hearing aids offer other useful features, such as noise suppression, which can filter out the background chatter in a crowded room without affecting the voice of the person to whom the wearer is listening. The difference between a custom hearing aid and what Aurilink is selling is akin to the difference between prescription glasses from an optometrist and a the $10 reading glasses you can buy at the local drug store.

Indeed, Aurilink’s CEO admits it:

“We like to think of the Sound Magnifiers as reading glasses for your ears,” said Otis A. Whitcomb, President and CEO of Aurilink.

That’s not to say that I think their product is useless. Far from it. Reading glasses are a vast improvement over no glasses at all. Aurilink’s products offer a similar value. I’m sure it’s an excellent product for what it does (even if it is built on our competitor’s platform, which is only a guess on my part), but “equal or superior to the most advanced DSP […] hearing instruments currently available?” Come on! Who are you kidding?

Canada Added to Piracy Watch List

According to The Globe and Mail, the U.S. has put Canada on a special watch list as part of its annual report on intellectual property rights. We share this honour with such illustrious nations as Ukraine, Belize, Latvia, Lithuania, Taiwan and Thailand.

“It is imperative that Canada improve its enforcement system so that it can stop the extensive trade in counterfeit and pirated products, as well as curb the amount of transshipped and transiting goods in Canada,” according to the report by the office of the U.S. Trade Representative.

The report also urged Canada to pass legislation to give its customs officers greater power to seize suspected pirated and counterfeit goods.

Welcome to the Axis of Piracy, folks.

Key to Successful Team Building

Luís Nunes Amaral at Northwestern University and his team have discovered one of the keys to building a successful team, according to a New Scientist story. After analysing the production teams responsible for Broadway musicals between 1877 and 1990, and scientific research teams from 1955 to 2004, they found that successful teams have mix of new and experienced collaborators:

The found they could predict success largely looking at just two parameters – the likelihood of a newcomer being in the team, and the likelihood of a collaboration being repeated. “We were very surprised because it worked. We were able to reproduce what was going on very nicely,” Amaral told New Scientist.

The researchers rated the success of scientific teams by examining the impact of the journals they published their work in, spanning ecology, astronomy, social psychology and economics.

“The teams publishing in good journals were built in a different way,” he says. Research teams publishing in lower impact journals tended to repeat collaborations again and again. The most successful teams did work with the same colleagues too, but only 75% of the time, he says.

One more reason it makes sense to hire co-op students.

When I worked at TI, we kept a small army of co-op students on staff every term. At one point, nearly half the engineering staff was students. Besides being cheap labour, they always brought a lot of fresh ideas with them. It made for a very energetic and fun workplace. I wondered at the time if it made the team more productive. I suspected it did, but it’s nice to see some research to back it up.

Solution to SICP Exercise 1.18

Structure and Interpretation of Computer Programs

Solution to Exercise 1.18:

(define (double x)
(+ x x))

(define (halve x)
(/ x 2))

(define (even? n)
(= (remainder n 2) 0))

(define (it-fast-mult a b)
(iter 0 a b))

(define (iter total a b)
(cond ((or (= a 0) (= b 0)) 0)
((= b 1) (+ total a))
((even? b) (iter total (double a) (halve b)))
(else (iter (+ total a) a (- b 1)))))

Solution to SICP Exercise 1.15

Structure and Interpretation of Computer Programs

Solution to Exercise 1.15:

Part a. Here’s the process of (sine 12.15).

(sine 12.15)
(p (sine 4.05))
(p (p (sine 1.35)))
(p (p (p (sine 0.45))))
(p (p (p (p (sine 0.15)))))
(p (p (p (p (p (sine 0.05))))))
(p (p (p (p (p 0.05)))))
...

There are clearly 5 applications of the p procedure.

Part b. We can see from the process above that the p procedure is applied for every power of 3.0 in angle. The order of growth in the number of steps is therefore Θ(log3n). Because the p procedures accumulate for every power of three, the growth of process in space is also Θ(log3n).