See this clip of a TV weather broadcast for yet another example of how American software generally can’t figure out how to set priorities.
“After ‘The Biggest Loser,’ Their Bodies Fought to Regain Weight” (nytimes) reports on a doctor who followed reality TV contestants and concluded that a slowed metabolism accounted for their inability to maintain weight loss.
I wonder if it should make us reconsider our notions of progress in medicine. In the 1950s an overweight person would have been diagnosed for $10 with a slow metabolism and given a $2/month amphetamine prescription to speed up said metabolism. Today the same person could be a $50,000/year customer for weight loss clinics, supplements, surgeries, etc.
Obviously being on amphetamine for years is harmful (see “America’s First Amphetamine Epidemic 1929–1971”), but is it as harmful as weighing 300 lbs?
Readers: What do you think? Should medicine concentrate on speeding up metabolism, possibly with new drugs, or is that also doomed due to the fact that people can just have a second donut?
A friend told me how pleased he was with MassHealth, our state’s version of Medicaid. “It is much better than Blue Cross. There are no deductibles and dental care is included and free,” said the father of two. His wife said “the only thing that would be better is if we got divorced and then I could get all of the single mom stuff.”
The potentially interesting part of this story is that my friend’s mailing address is a suburban house on 2.6 acres of land with a Zestimate of $3.2 million and a price/tax history indicating that he purchased it for $2.1 million back in 2006. The health insurance ministry probably wouldn’t have been aware of the family’s 50-acre 8-bedroom (including guest house) vacation estate.
“Obamacare removed the asset test for Medicaid,” he explained. So they look only at income? “No. I actually had a good year in 2015 with a lot of capital gains.” [“A lot” of capital gains for this guy would be hundreds of thousands or single-digit millions of dollars.] What was the method for determining whether or not the taxpayers would pick up the tab for this family of four’s health and dental care? “They look only at W-2 income.”
Note that this makes collecting alimony and child support relatively more lucrative compared to working. Child support revenue, regardless of the amount, doesn’t count as “income” to qualify for this taxpayer-funded benefit. Alimony profits can be banked without a W-2 being issued. Jessica Kosow, the plaintiff in a typical higher-income Massachusetts case (see this chapter on Massachusetts family law), would qualify for free MassHealth despite having obtained, via litigation, triple the spending power of her Ivy League classmates with W-2 jobs:
In a June 22, 2011 status conference for this case [wife sued husband after four years of marriage when their daughter was two years old], Judge Maureen Monks explained her philosophy in setting child support for high income defendants: “when I look at how the current guidelines play out against most parties’ income it comes around between 20 and 25 percent, sometimes it’s a little higher. If there’s a big disparity it’s closer to 28 percent. Does that mean it makes sense is that what to assess up to a certain amount on his income. Maybe there is no limit right now…”
How did she do compared to her University of Pennsylvania classmates? payscale.com reported that in 2014 the median “mid-career” salary for a graduate of this Ivy League college was $112,200. If that graduate stayed in Pennsylvania, his or her earnings would be approximately $77,240 per year after taxes (ADP Paycheck Calculator). Kosow’s after-tax earnings, on the other hand would be approximately $132,786 in cash plus the free $1 million house (assume a rental value of $6000 per month), health insurance, and nanny services. Her total after-tax earnings from the Massachusetts divorce and child support system therefore would be about $250,000 per year, 3.2X what a Penn graduate working full-time would earn.
[Of course, this particular plaintiff might not take the time to fill out the forms for MassHealth due to the fact that she obtained a court order that her former husband pay her health insurance bill.]
How about this for an AI project: a computer system that takes the audio stream of a National Public Radio station and presents a commercial-free audio stream as an output, presumably with at least a one-minute delay.
It should be pretty easy for a classical music station. There isn’t too much talking and the non-commercial stuff is typically right before or after a piece. Even a program that simply muted humans talking and let through music would perhaps be sufficient (though watch out for Recitative!).
For an NPR talk/news station it would presumably be tougher because the same announcers present both content and commercials. However, that’s what makes it a great project for a Master’s student in AI if not a PhD.
What do readers think? Should a prize be established for the best programs that can do this? It seems like a useful application of AI and also something where it is easy to test and score programs (sum of minutes of content improperly muted and minutes of commercials improperly not muted, for example). Programmers around the world could compete for honors/money.
I was recently scheduled to be deposed as an expert witness in a software patent case. The Stuffy Big Firm (founded in the 19th century; more than 1000 lawyers today; breathtaking office space) partner representing the defendant said that the deposition would be videotaped. I said “Oh, so I’ll need to wear a suit.” He said “Don’t bother. Nobody will ever look at it.”
“The Voyeur’s Motel” is a New Yorker story mostly about sex. But the subject of this blog posting is a section buried in the middle:
[The voyeur, who ran a motel in Colorado,] also got bored with cataloguing his guests’ dishonesty. They sometimes tried to cheat him out of the room rent, and hardly a week passed without his witnessing instances of chicanery. One working-class couple asked him for a few days’ grace period to pay their bill. Foos spied on them the next day and heard the husband tell the wife, “The dumb guy in the office thinks I have a check coming in from Chicago, and we will fool him the same way we did the motel in Omaha.” Foos locked the people out of their room and kept their possessions until they paid him.
Diary excerpt: “Conclusion: Thousands of unhappy, discontented people are moving to Colorado in order to fulfill that deep yearning in their soul, hoping to improve their way of life, and arrive here without any money and discover only despair. . . . Society has taught us to lie, steal, and cheat, and deception is the paramount prerequisite in man’s makeup. . . . As my observation of people approaches the fifth year, I am beginning to become pessimistic as to the direction our society is heading, and feel myself becoming more depressed as I determine the futility of it all.”
These experiences prodded Foos to concoct an “honesty test.” He would leave a suitcase, secured with a cheap padlock, in the closet of a motel room. When a guest checked in, he would say to Donna, in the guest’s hearing, that someone had just called to report leaving behind a suitcase with a thousand dollars inside. Foos then watched from the attic as the new guest found the suitcase and deliberated over whether to break the lock and look inside or return the suitcase to the motel office.
Out of fifteen guests who were subjected to the honesty test, including a minister, a lawyer, and an Army lieutenant colonel, only two returned the suitcase to the office with the padlock intact. The others all opened the suitcase and then tried to dispose of it in different ways. The minister pushed the suitcase out the bathroom window into the bushes.
A lot of U.S. government programs are set up with the idea that Americans are fundamentally honest. Offering enhanced payments for disability is not going to change the number of people who seek to collect SSDI (see “Book Review: The Redistribution Recession“). Certainly we wouldn’t ever expect 97 percent of retired government workers to claim disability benefits. Hardly anyone would have sex with a drunken married dentist in order to harvest the $millions in tax-free child support that a Massachusetts or Wisconsin court would hand out. Nobody would work the FMLA to get full-time benefits out of a part-time job. People aren’t going to work in cash jobs in order to remain eligible for free public housing.
Readers: What do you think? Do we trust this motel owner’s data? If the data are right, is much of the current U.S. system set up improperly?
The New York Times didn’t wait to hear the evidence before lauding as meritorious Ellen Pao‘s employment discrimination claim against the one-percenters at Kleiner Perkins. Now they’ve got their own plaintiff, Marjorie Walker (“Here’s why I’m suing the New York Times for discrimination” (Guardian)), alleging that they discriminate against “older, minority and female employees” in favor of “young, high-end and primarily white” workers.
One common thread is that the Times doesn’t have any difficulty in making up its mind about these lawsuits. Just as they were quickly sure that Ellen Pao was right, they’re confident that Marjorie Walker is dead wrong:
A spokeswoman for The Times called the suit “entirely without merit” and said “we intend to fight it vigorously in court.’’ — “Suit Accuses New York Times Executives of Bias”
What do readers think? Ellen Pao’s argument was that the Kleiner Perkins partners wanted to make themselves poorer by bringing in a less qualified person as a partner (my initial posting on the lawsuit where I wondered what their motivation could be). Marjorie Walker and her co-plaintiff Ernestine Grant argue that mid-level managers at the Times are favoring workers under 40 rather than workers over 60.
- “Sheryl Sandberg, Jane Austen, and the Queen of Versailles” — in the middle is another story that the Times wasn’t enthusiastic about covering; one of its reporters turned an adulterous affair into child support profits worth roughly 100 years of income for a journalist (a fairly straightforward operation under New York family law)
- “It’s Silicon Valley 2, Ellen Pao 0: Fighter of Sexism Is Out at Reddit”, a post-lawsuit Times piece in which Ellen Pao is characterized as a “fighter” against sexism that the half-female jury found did not exist.
- “Ellen Pao Suit Against Kleiner Perkins Heads to Trial” is a typical pre-lawsuit piece that leads with a paragraph about how “many women in technology … are rarely hired, promoted or taken seriously, and are confronted on a daily basis by sexism and harassment. They feel demeaned and discouraged.” Ellen Pao having sex with a married partner is described with “Ms. Pao says a married colleague pressured her into an affair”. Kleiner Perkins’s defense is “a scorched-earth response.”
- “Lawsuit Shakes Man’s World of Tech” was an early Times piece stripped down to “even some of Ms. Pao’s critics concede that she is exposing an uncomfortable truth about Silicon Valley” (sadly, not the “truth” was not true enough for the jury to give Pao the $160 million she was seeking in her lawsuit (compare to $227 million she could have gotten for having sex with her boss))
An evergreen subject for our media seems to be “female” athletes earning less than athletes who identify as “male”. “Roger Federer, $731,000; Serena Williams, $495,000: The Pay Gap in Tennis” is a recent example from the New York Times.
I’m wondering if games would be more interesting if there were no restrictions on gender identification. Certainly right now the singles game is typically pretty predictable and often the winner is simply the taller player. Could an amazing player 5’8″ tall reasonably hope to beat Novak Djokovic at 6’2″?
Why not deal squarely with the issue that huge strong players, regardless of gender ID, can overpower physically smaller players? We can group sub-tournaments by weight and/or serve speed rather than by gender ID. If we are concerned about pay equity then just have the prize money be the same in each group.
[Separately, what is there to stop a 6’4″-tall “male” player right now from putting on a skirt, saying “I identify as a woman”, and entering a woman’s tournament? Gael Monfils, for example, probably wouldn’t have too much trouble overpowering even a top player currently identifying as “Female”.]
“MIT built its own Ellen Pao before the Ivy League did: Gretchen Kalonji” covers a UC chancellor who arranged a sinecure for her lover and ran up $600,000 in renovation bills before it all ended in suicide and litigation.
Today’s nytimes has “University of California, Davis, Chancellor Is Removed From Post” about Linda Katehi and “questions about the campus’s employment and compensation of some of the chancellor’s immediate family members.”
Do they really need to have someone in this position if it is so prone to nepotism?
People seem to be excited about government regulations around self-driving cars. I’m wondering why this couldn’t be reduced to “Your self-driving car needs a $20 million liability insurance policy from an A-rated company.” The insurers aren’t going to underwrite high risks, presumably.
On a boat tour of Ft. Lauderdale’s rich waterfront houses, the captain/guide told us of his experience working on big private yachts. I asked him “If I am a rich douchebag can I buy a crazy long yacht tomorrow and, knowing nothing about boats, start driving it around?” He explained that Coast Guard regulations are essentially irrelevant because long before a regulation would require training or certification the boat’s insurer would insist on a licensed captain and perhaps additional crew members.
This is also true in aviation. A woman who got her pilot certificate yesterday can get into a $5 million Pilatus PC-12 and push the start button from an FAA point of view. But no insurance company would allow that.
Why not assume that underwriters are prudent and can evaluate the risks of self-driving vehicles at least as well as government workers and politicians?
[Maybe tweak this with a requirement also for comprehensive data and video logging in any self-driving car so that it is easier to determine if the self-driving vehicle caused an accident. But that would still be only a couple of paragraphs of regulation.]