Seeing at the Speed of Sound

That's the title of a lovely piece by Stanford graduate Rachel Kolb about what it's like for a deaf person to try to lipread.

Lipreading, on which I rely for most social interaction, is an inherently tenuous mode of communication. It's essentially a skill of trying to grasp with one sense the information that was intended for another. When I watch people's lips, I am trying to learn something about sound when the eyes were not meant to hear.

Spoken words occur in my blind spot, a vacancy of my perception. But if I watch a certain way, I can bring them into enough focus to guess what they are. The brain, crafty as it is, fills in the missing information from my store of knowledge.

The article gave me both a deep empathy for what it is like to try to lipread, how complex an art it is (at best, Kolb says she understands 30% of what she tries to lipread), and also a great appreciation for what it must be like when she is able to "see" the language come into sharp focus on another person's lips.

SOME PEOPLE ARE all but impossible for me to lipread. People with thin lips; people who mumble; people who speak from the back of their throats; people with dead-fish, unexpressive faces; people who talk too fast; people who laugh a lot; tired people who slur their words; children with high, babyish voices; men with moustaches or beards; people with any sort of accent.

Accents are a visible tang on people's lips. Witnessing someone with an accent is like taking a sip of clear water only to find it tainted with something else. I startle and leap to attention. As I explore the strange taste, my brain puzzles itself trying to pinpoint exactly what it is and how I should respond. I dive into the unfamiliar contortions of the lips, trying to push my way to some intelligible meaning. Accented words pull against the gravity of my experience; like slime-glossed fish, they wriggle and leap out of my hands. Staring down at my fingers' muddy residue, my only choice is to shrug and cast out my line again.

Some people, though not inherently difficult to understand, make themselves that way. By viewing lipreading as a mysterious and complicated thing, they make the process harder. They over-enunciate, which distorts the lips like a funhouse mirror. Lips are naturally beautiful, especially when words float from them without thought; they ought never be contorted in this way. There are other signs, too: nervous gestures and exaggerated expressions, improvised sign language, a tic-like degree of smiling and nodding.

The World's Top NBA Gambler

A fascinating profile in ESPN Magazine of Bob Voulgaris: Meet the world's top NBA gambler​. Together with a math, statistics, and programming prodigy Voulgaris simply calls the Whiz (he won't reveal the Whiz's real identity for fear of having him poached), Voulgaris built an NBA simulator named Ewing, after Bill Simmons' Ewing Theory.

If Ewing has a secret sauce, it’s just this sort of thing: Finding scraps of information, sliced and diced ever more finely, that reveal something about how a system -- in this case, a game of pro basketball -- will operate in the future. The key is to find those scraps that are more predictive than others. Case in point: One of Ewing’s most important functions is to assign values to players. Each player has two values -- on offense and as a defender -- and those values are constantly changing. Ewing will also automatically adjust the value depending on who’s guarding whom. Oklahoma City’s Kendrick Perkins “is more valuable guarding Dwight Howard than he is guarding Shane Battier,” Voulgaris says. Why? “Because Howard is a unique player, and you need a big to defend him.” Likewise, according to Voulgaris, Celtics seven-footer Jason Collins is “useless every game, except when he’s guarding Howard, which he does really, really well.” Player values also change across a season and a career. So Voulgaris and the Whiz created, for Ewing, an aging component. Further number-crunching revealed that different types of players, based on position and size, will reach their zeniths at different ages and on trajectories that are possible to predict. Ewing now grasps the curve of the lifespan of the point guard, the shooting guard, the forwards, the center -- and predicts the downslope and expiration date of every NBA career. 

When Ewing went live with actual betting for the first time toward the end of the 2008 season, Voulgaris was not yet sold on its powers. For one thing, his subjective-gambler side wasn’t ready to surrender control to a machine. For another, the model was performing unremarkably with their money on the line -- right above the break-even line. But Voulgaris had something in mind, “a long project, like a six-month-long project, to model a certain part of the game of basketball.” He and the Whiz spent the offseason pursuing this mysterious project, the precise nature of which Voulgaris will not discuss. “I don’t even want to allude to what it might be,” he says when I press him, “because I don’t think anyone else is doing anything like it.” 

By 2009, once they’d added this mysterious additional model to Ewing’s inner workings -- version 2.0 -- they started making bets based on the scores it produced after the All-Star break. “We just, like, crushed the second half of the season,” Voulgaris says. Since then, as each subsequent season has passed, Voulgaris’ confidence in Ewing has increased. So too has the frequency of his wagering. In a season, he now regularly puts down well over 1,000 individual bets. “I mean, I don’t want to sit here and brag,” he says. “But this is literally, like, the greatest thing ever when it comes to sports betting.” 

​More money is gambled on the NFL than any of the major US sports, but given how strong a role luck plays in NFL outcomes, it's surprising more people don't gamble on the NBA instead since skill plays a greater role in the NBA than in MLB, the NBA, or the NHL.

The singularity already happened

One of my favourite recurring tropes of AI speculation/singulatarian deep time thinking is mediations on how an evil AI or similar might destroy us.

...

And all I can think is: we already have one of those. It is pretty clear to anyone who’s paying attention that 1. a marketplace regime of firms dedicated to maximizing profit has—broadly speaking—added a lot of value to the world 2. there are a lot of important cases where corporate profit maximization causes harm to humans 3. corporations are—broadly speaking—really good at ensuring that their needs are met.

From Mini. Quiet Babylon.​ Intriguing.

Some of this is a time horizon issue. ​At some sufficiently long time horizon, things like environmental destruction or, to take a more popular AI trope, the destruction of the human race, would seem to be counter to a profit motive.

Perhaps that's the wrinkle in this hypothesis. For a variety of reasons, corporations have failed to survive for very long; most have a lifespan shorter than that of the average human (the open market being a ruthless jungle to try to survive in). Not long enough, I suspect, to execute on a long-term plan to destroy humanity. Given this theory, we should perhaps regard antitrust regulation with even more respect.

The linked post mentions that corporations have corrupted American politics. Jonathan Rauch makes a great argument in his book Government's End why the American political system is hugely susceptible to the influence of special interests, and it's not just the profit motive but inherent structural flaws in the American (and many other) forms of government.

Alan published in NEJM

My brother Alan had an article published in the February 14th issue of The New England Journal of Medicine. We're all proud of him.​

The article title: Selumetinib-Enhanced Radioiodine Uptake in Advanced Thyroid Cancer.​ As with all brilliant ideas, the conclusion of the article seem self-evident upon further reflection, I mean clearly you'd anticipate selumetinib producing clinically meaningful increases in iodine uptake and retention in a subgroup of patients with thyroid cancer that is refractory to radioiodine, it's amazing we never believed this before, ahhh, I have no idea what I'm talking about, why am I not smart enough, my life has no meaning.

A dangerous market feedback loop

Moskowitz’s path to mastering the bliss point began in earnest not at Harvard but a few months after graduation, 16 miles from Cambridge, in the town of Natick, where the U.S. Army hired him to work in its research labs. The military has long been in a peculiar bind when it comes to food: how to get soldiers to eat more rations when they are in the field. They know that over time, soldiers would gradually find their meals-ready-to-eat so boring that they would toss them away, half-eaten, and not get all the calories they needed. But what was causing this M.R.E.-fatigue was a mystery. “So I started asking soldiers how frequently they would like to eat this or that, trying to figure out which products they would find boring,” Moskowitz said. The answers he got were inconsistent. “They liked flavorful foods like turkey tetrazzini, but only at first; they quickly grew tired of them. On the other hand, mundane foods like white bread would never get them too excited, but they could eat lots and lots of it without feeling they’d had enough.”

This contradiction is known as “sensory-specific satiety.” In lay terms, it is the tendency for big, distinct flavors to overwhelm the brain, which responds by depressing your desire to have more. Sensory-specific satiety also became a guiding principle for the processed-food industry. The biggest hits — be they Coca-Cola or Doritos — owe their success to complex formulas that pique the taste buds enough to be alluring but don’t have a distinct, overriding single flavor that tells the brain to stop eating.

​From the tomorrow's NYTimes Magazine cover story The Extraordinary Science of Junk Food. It's both fascinating and terrifying.

Poring over data one day in his home office, trying to understand just who was consuming all the snack food, Riskey realized that he and his colleagues had been misreading things all along. They had been measuring the snacking habits of different age groups and were seeing what they expected to see, that older consumers ate less than those in their 20s. But what they weren’t measuring, Riskey realized, is how those snacking habits of the boomers compared to themselves when they were in their 20s. When he called up a new set of sales data and performed what’s called a cohort study, following a single group over time, a far more encouraging picture — for Frito-Lay, anyway — emerged. The baby boomers were not eating fewer salty snacks as they aged. “In fact, as those people aged, their consumption of all those segments — the cookies, the crackers, the candy, the chips — was going up,” Riskey said. “They were not only eating what they ate when they were younger, they were eating more of it.” In fact, everyone in the country, on average, was eating more salty snacks than they used to. The rate of consumption was edging up about one-third of a pound every year, with the average intake of snacks like chips and cheese crackers pushing past 12 pounds a year.

Riskey had a theory about what caused this surge: Eating real meals had become a thing of the past. Baby boomers, especially, seemed to have greatly cut down on regular meals. They were skipping breakfast when they had early-morning meetings. They skipped lunch when they then needed to catch up on work because of those meetings. They skipped dinner when their kids stayed out late or grew up and moved out of the house. And when they skipped these meals, they replaced them with snacks. “We looked at this behavior, and said, ‘Oh, my gosh, people were skipping meals right and left,’ ” Riskey told me. “It was amazing.” This led to the next realization, that baby boomers did not represent “a category that is mature, with no growth. This is a category that has huge growth potential.”

​The article includes wonderful tidbits like "people like a chip that snaps with about four pounds of pressure per square inch" and explains why Cheetos are one of the most perfect snacks ever constructed.

The foodie movement look to high end restaurants ​for culinary innovation, but the truth is that much more of that happens in the mass market industrial food production machine. Many high end restaurant techniques are actually borrowed from the industrial food production laboratories.

This all speaks to one of the defects of our free market economy, that these dangerous feedback loops will be set up in which we are given exactly what we want but don't need. The most insidious type of killing might be the one that happens under our very noses, so slowly we don't notice it, a caper in which we are given cheap and ready access to a slow-acting poiso and readily gorge on it until it's too late.

Reading about some of these brilliant food scientists, concocting new snacks to steal our market share, I couldn't help but think of Walter White, with his blue meth. In the tech industry, it's fashionable to talk about marketing and distribution as necessary companions to product development. Few industries embody the perfect unity of those disciplines than the food industry.