Month: September 2015

I recently visited Virginia Tech (note how Google shows Nidal Malik Hasan as #1 among the school’s “notable alumni,” a good reason to invest in PR…).

The school impresses the visitor with its NFL-grade football stadium, Four Seasons-grade hotel and conference center, and stone buildings (but the architecture school is a hideous concrete eyesore! (photos)). Due to the handsome stone construction, the school has some of the feel of Princeton, generally admired as one of America’s nicest college campuses. With 31,000 students Virginia Tech needs to be physically larger than Princeton, with 8,000 students. However, the physical inflation is more dramatic than the student ratio.

My theory is that Virginia Tech is spread out to a large extent because of the need to provide massive surface parking lots in front of most buildings, a requirement dictated by (a) the 100-year period of private ownership of non-self-driving cars, and (b) the expense of digging underground lots. It is tough to imagine casual collaboration among departments at Virginia Tech due to the long walks from subcampus to subcampus. Because the campus isn’t all that walkable, there is a huge amount of car traffic even after 9 pm on weekdays.

I’m wondering if once the self-driving (and shared) car (or RV) is here we will regret having built so many spread-out campuses to accommodate what turned out to be a relatively small fraction of the period of European settlement of the Americas.

Full post, including comments

http://www.teslamotors.com/modelx will shortly show the launch of the new electric SUV (not “pavement-melting SUV,” which had become almost a Homeric epithet). Fully loaded this is a $132,000 car. If you pay federal income tax you will be contributing to a $7,500 subsidy for each person (a.k.a. “rich bastard”) who buys one. It seems like a marvelous engineering achievement so… who else is excited to be offered the opportunity to work extra hours every year in order to help rich people buy this new toy?

And when do we get to subsidize an all-electric Rolls-Royce?

[Disclosure: A rich friend of mine has one of these on order and I am looking forward to riding in his.]

Full post, including comments

Paul McHugh, a retired psychiatrist from Johns Hopkins, summarizes the state of modern psychiatry in a review of Shrinks, a book that tries to do the same thing. A friend whose PhD is in experimental psychology (Stanford) sometimes irritates therapists by asking “How do you do engineering when there is no physics?” McHugh makes the same point:

No formative principle or principles link the function of the brain to mental disorders, and until such is discerned, psychiatry will remain at the naming and describing level…

The end of the beginning for cardiology came when William Harvey wrote De Motu Cordis in 1628. Therein he described how the heart acted as a pump, how the blood circulated, and how the experimental method could advance human physiology. Everything in cardiology—from understanding a patient’s symptoms to discovering mechanism-based medications, right on to transplanting hearts today—rests on that foundation.

Psychiatry has no Harvey. And if DSM-5 is its “Bible,” then the “end of the beginning” for psychiatry looks decades away.

Why is psychotherapy so popular? A Harvard PhD psychologist said that “sociopaths and psychopaths love to go to therapists because part of the standard process is unconditional positive regard. They’re not going to get that from anyone whom they talk to and aren’t paying.”

What do readers think? What kinds of science would we need to enable therapists to be reliably effective? And where will that science come from? Neuropsychology? Or perhaps from the more heavily funded Artificial Intelligence folks? Are we talking about 10 years or 100 years before we see significant progress?

Separately, what success stories for psychotherapy do readers have to relate? Does anyone have personal experience with someone who went to therapy and came out as a substantially improved individual? (if drugs, such as Prozac, were involved, please limit your answer to situations in which you’ve had at least five years to observe the results (to reduce distortions from the placebo effect))

Full post, including comments

“Gender equality: Taking stock of where we are” is an article from McKinsey, management consultant to Enron (Guardian). This cluster of Harvard and Stanford MBAs has figured out that setting explicit quotas results in more employment for the targeted group:

Our experience has been that top-down targets make a difference. We didn’t set explicit gender goals for McKinsey until 2014, and in just one year after doing so, our intake of female consultants has increased by five percentage points.

McKinsey charges customers a lot of money for forward thinking, but they seem to be taking a backward attitude toward gender. What does it mean for a consultant to be “female”? How does McKinsey know that 100 percent of their employees won’t come in tomorrow morning identifying as “female” (or “male” for that matter)? If the modern way of looking at gender is that it is primarily a state of mind, wouldn’t the smartest employers achieve instant gender balance by offering a bonus to any worker of the over-represented gender to identify as a member of the under-represented gender? (See this report on papers from the American Economic Association on what people are willing to do in exchange for small financial incentives.)

Let’s assume for the moment that McKinsey is correct in its assumption that employee gender is persistent and/or biological. Why is their focus on the gender composition of their workforce? Does McKinsey have the same percentage of African-American partners in the U.S. as there are African-Americans in the U.S. population as a whole? If not, does switching the focus to gender mean that McKinsey has given up on achieving racial goals/quotas? (Or that McKinsey agrees with Rachel Dolezal regarding race being a social construction?) One thing that nearly everyone agrees on, except perhaps Tinder users, is that age is an immutable biological fact for humans. Why isn’t McKinsey interested in addressing age discrimination, establishing hiring quotas for older workers, or examining the potential for age disparity in new hires that results from recruiting almost exclusively from among fresh MBA graduates?

Let’s assume that the gender composition of a workforce is in fact something that should take higher priority than racial or age composition. According to the authors, McKinsey is interested in sorting employees by gender so that women can have “economic equality”:

Economic equality for women, to no small degree, depends on achieving a sweeping set of social-equality reforms. Is it the business of executives to help solve broader social issues? We would say yes, provided they don’t distract from the very real issues executives face in their own organizations.

The article does not mention, however, that women in many parts of the U.S. can obtain the after-tax spending power of a McKinsey partner by having sex with three McKinsey partners and harvesting the resulting child support (see Real World Divorce; works best in California, Massachusetts, New York, and Wisconsin; if she doesn’t want children she could have sex with six McKinsey partners and sell the abortions). If a woman and a man have the same spending power, isn’t that “economic equality”? If the goal is economic equality, shouldn’t McKinsey therefore make greater efforts to achieve gender balance in jurisdictions, e.g., Germany and Scandinavia, where unlimited child support profits are not available?

McKinsey is a member of the “30% club,” seeking a quota of 30% for corporate board representation, but the authors don’t explain why the goal should be 30 percent. Why not 50 percent, for example? Or 100 percent for a period of time in order to make up for past underrepresentation? (And if McKinsey has denied employment opportunities to women in the past, doesn’t fairness require that the quota for new hires be 100% female, at least for a period of time?)

The authors claim that it will be tough for McKinsey to hire and retain more women but they don’t explain why. Is it the case that among the 7 billion people on Planet Earth there are not enough qualified people who identify as “women” to occupy 50, or even 100 percent of the chairs at McKinsey? If there are sufficient qualified people who, at least for the present identify as “women,” what is preventing McKinsey from offering them a combination of salary and benefits that would induce them to join and stay at McKinsey?

If McKinsey is not competent to hire the mix of employees that it desires, why announce that failure to the world? Doesn’t that cast some doubt on McKinsey’s competence to serve as management consultants? Or is McKinsey’s point that they ran a quota system at their company, which bumped up the percentage of employees who identify as part of the female gender, and you can pay them to advise on running a similar quota system at your company?

What do readers think? Does this article make you more or less impressed with the quality of thinking at McKinsey?

Full post, including comments

A friend with four children has spent perhaps 400 hours repeating the same order: sit down in your chair. Why? If a child is standing up and grasping the back of the chair there is a risk of tipping over (static rollover).

Now that we find it necessary in our own household to replay these scenes I am wondering if there isn’t a technical solution. Why not kitchen chairs with insanely heavy lead-filled legs? Then the toddlers can do whatever they want on the chairs without risk of toppling. Probably have to combine the lead-filled legs with appropriate geometry (legs contact the ground behind the chair back) to avoid a dynamic rollover.

Has anyone tried this?

Separately, what the most tip-proof chairs that are easily found on the U.S. market right now? Could it be that standard office side chairs (Allsteel Relate Side as an example) are the most stable? The back legs seem to go pretty far back, as though the designers were trying to ensure that the cubicle farm remained a safe zone for all sitting styles.

Related:

• untippable chair from England (but designed for students sitting in the chair, not standing on it)

Full post, including comments

Today is my 52nd birthday, an unpleasant reminder that at some point in the (nearer) future I can expect to become more reliant on the medical profession.

If we consider doctors as information processing systems (data in, diagnosis out), how are they improving compared to other information processing systems? “Medicine’s Uncomfortable Relationship With Math,” a 2014 paper from JAMA Internal Medicine shows no improvement over the past 30 years. 75 percent of the docs surveyed couldn’t figure out how to use data from a test for a rare condition: “Of 61 respondents, 14 provided the correct answer of 2% [likelihood that a patient was actually suffering from the condition, based on having tested positive]. The most common answer was 95%, provided by 27 of 61 respondents. The median answer was 66%, which is 33 times larger than the true answer.”

 

Full post, including comments

“BBA Aviation Buys U.S. Rival Landmark for $2 Billion” (WSJ) explains that if you want jet fuel at the busiest airports you’ll now have one source, Signature, instead of two (Signature and Landmark). Why wouldn’t the antitrust regulators whose salaries and benefits are funded by our tax dollars object to this reduction in competition? “Mr. Pryce said the market remains highly fragmented and that he didn’t expect regulatory hurdles.”

Is the market truly fragmented? Well… if you’re going from Boston to New York City and you want to stop for fuel in rural Pennsylvania you can get a great deal on jet fuel from an independently owned and/or municipal supplier. If you want to land and fuel at, for example, Westchester, however, you’ll see that there are currently just three choices: Landmark, Million Air, and Signature. Due to current vibrant competition, the listed price on Airnav is about $7.40 per gallon at Landmark and Signature. Compare to nearby Danbury, where the runway is a little too short for Landmark and Signature to be interested: $3.50/gallon. In other words, they’re already able to command insane profits at Westchester.

This merger makes as much sense to me as letting Comcast and Time-Warner join forces… (previous posting) Why is it that so few Americans are interested in the “competition” requirement that makes the rest of Econ 101 potentially relevant? Typical Americans don’t own Gulfstreams, of course, but Americans who own shares in public companies or who buy insurance from mutually owned insurance companies do pay for jet fuel to go into Gulfstreams.

[Separately, Signature is owned by a foreign company. So this means that the profits that Signature obtains by limiting competition will be sent over to England.]

Full post, including comments

… because for every immigrant a job at the FBI will be created.

See “The Imam’s Curse,” a New Yorker magazine article about a family of Pakistani immigrants who have generated FBI jobs at a 1:1 ratio.

[Note that these folks would not have been able to emigrate to Australia or New Zealand under those countries’ “point” systems. Their lack of education and, in the Imam’s case, old age, would have eliminated them from consideration.]

Full post, including comments

One relatively new feature of the American intellectual property system is the “covered business method review” procedure run by the USPTO. Pretty much everything that happens in front of the U.S. Patent Trial and Appeals Board (PTAB) is public and available from ptabtrials.uspto.gov. A case in which I became involved as an expert witness in April 2014 was recently decided. I thought that readers might be interested to learn about the process. If you type in Case No. “CBM2014-00125” (or 126, 127, 128 to see the dockets for the related patents) at the ptabtrials site you can find everything (about 140 documents per case, many duplicated and/or with duplicated sections).

The dispute began in December 2013 when Paid, Inc. sued eBay for infringing four patents on calculating shipping costs (example patent; amended complaint). In parallel with defending the lawsuit in federal court, eBay in May 2014 filed four petitions for covered business method review (example; each CBM review has its own parallel set of paperwork). Each included a declaration from me (example). Paid, Inc. responded in August 2014 (example). In September 2014 the PTAB decided to institute review of the four patents (example). Paid, Inc. responded to that in January 2015 (example) with a single declaration from its own expert (link) and one from an inventor (link). eBay’s attorneys deposed the expert (transcript) and inventor (transcript) and filed a reply in April 2015, with a supplemental declaration from me. An oral hearing at which attorneys argued was held in June 2015 (transcript). A final decision was issued in September 2015.

Note that this is not necessarily the end of the legal road. Decisions of the PTAB that are adverse to patent owners can be appealed to a Federal Circuit Court.

[Sidenote/Opportunities Missed Department: In 1995 I was working for a large publisher and had the following conversation with management:

• Me: “You guys own newspapers all around the country with classified ad sections. How about giving classified ad buyers a free web listing and having a server aggregating classified ads from all around the nation?”
• Them: “How would that work?”
• Me: “Classified ads are pretty much free to display on the Web so you might want to look at the readers as giving us valuable content alongside which we can sell display ads. Once the ad is on the web we can let people offer the item at a fixed price or let the server run an auction for them. Because people won’t always be buying locally we can have a system that lets buyers and sellers report on or review each other.”
• Them: “That doesn’t sound very interesting.”
• Me: “I actually already built a working prototype. Do you want to see a demo?”
• Them: “Sure.”
• *** Demo ensued ***
• Them: “Still not something we’re interested in doing.”
• Me: “Do you mind if I take the code and run it on my photo.net site?”
• Them: “We don’t care. You can do whatever you want with it.”

Thus was born the photo.net classifieds, described in this chapter of Database-backed Web Sites. (You might ask why I built a complete functional system before getting the green light from management. This was before the era of CSS and JavaScript and thus I was able to build the whole system in three days.)]

Full post, including comments

Our flight school went through a period of hiring Norwegian citizens as helicopter instructors. These guys went through a rigorous training program here in the U.S. and then had the right to work for 12-18 months as “practical training.” After that, the U.S. immigration bureaucracy worked aggressively to push these skilled English-fluent workers out of our country in favor of more folks such as Dzhokhar Tsarnaev and his family.

I caught up with one of these alumni recently and found out that he was flying the magnificent Sikorsky S-92 helicopter out to offshore oil rigs. At the current exchange rate, which is much less favorable for Norwegians than previously, a first officer earns about $100,000 per year and a captain $150,000 per year. They work 14 days on, 14 days off. The “on” days consist of flying 6-8 hours while the off days can be spent in Oslo or anywhere else in the world. Housing near the base costs about $750 per month and is the responsibility of the pilot. Commercial 45-minute flights to or from Oslo cost about $100 (another $200 per month).

I learned that finding the offshore rig is not the idiot-proof process that one would imagine. Instead of a GPS-guided approach down to the helipad the pilots look for the rigs using the S-92’s weather radar. If they can’t identify the rig visually within 0.75 nautical miles then it is time for a missed approach. Many offshore rigs have a similar appearance and are clustered together. Therefore it is not uncommon for a helicopter to land on the wrong rig.

Related:

 

• sad S-92 crash on the way out to a Canadian rig
• Lockheed Martin buys Sikorsky for 1.1X revenue

Full post, including comments