Lunch at the Labs

Various web links I found to be of interest recently (although a couple are older).

• Discovering That Denial of Paralysis Is Not Just a Problem of the Mind

  Sandra Blakeslee. The New York Times, August 2, 2005

  Dr. Anna Berti sits facing a patient whose paralyzed left arm rests in her lap next to her good right arm. "Can you raise your left arm?" Dr. Berti asks.
  "Yes," the patient says.
  The arm remains motionless. Dr. Berti tries again. "Are you raising your left arm?" she asks.
  "Yes," the patient says. But the arm still does not move.
  Dr. Berti, a neuroscientist at University of Turin in Italy, has had many such conversations with stroke patients who suffer from denial syndrome, a strange disorder in which paralyzed patients vehemently insist that they are not paralyzed.
  This denial, Dr. Berti said, was long thought to be purely a psychological problem. "It was a reaction to a stroke: I am paralyzed, it is so horrible, I will deny it," she said.
  But in a new study, Dr. Berti and her colleagues have shown that denial is not a problem of the mind. Rather, it is a neurological condition that occurs when specific brain regions are knocked out by a stroke.
  Patients deny the paralysis because a closely related region of the brain that is still intact appears to tell them that their bodies are responding normally.

• Robots Are Really Bad At Folding Towels

  All Things Considered. May 19, 2015

  In other words, years of work from dedicated, smart researchers have produced a towel-folding robot that can't keep up with an average 8-year-old. This problem, Abbeel says, is not limited to towels.

• Why Mastering Language Is So Difficult for AI

  Interview with Gary Marcus

  Marcus, who earned his Ph.D. in brain and cognitive sciences from MIT and is now a professor emeritus at New York University, says the field of AI has been over-reliant on deep learning, which he believes has inherent limitations. We'll get further, he says, by using not only deep learning but also more traditional symbol-based approaches to AI, in which computers encode human knowledge through symbolic representations (which in fact was the dominant approach during the early decades of AI research).
  ...
  Just as GPT-3 doesn't really understand language, merely memorizing a lot of traffic situations that you've seen doesn't convey what you really need to understand about the world in order to drive well. And so, what people have been trying to do is to collect more and more data. But they're only making small incremental progress doing that. And as you say, there aren't fleets of self-driving taxis in Toronto, and there certainly aren't fleets in Mumbai. Most of this work right now is done in places with good weather and reasonably organized traffic, that's not as chaotic. The current systems, if you put them in Mumbai, wouldn't even understand what a rickshaw is. So they'd be in real trouble, from square one.

  Also see Gary Marcus articles at Deep Learning Is Hitting a Wall and in Scientific American: Artificial General Intelligence Is Not as Imminent as You Might Think

• The race to invent new particles is pointless

  Sabine Hossenfelder: but no one in physics dares say so

  It has become common among physicists to invent new particles for which there is no evidence, publish papers about them, write more papers about these particles’ properties, and demand the hypothesis be experimentally tested. Many of these tests have actually been done, and more are being commissioned as we speak. It is wasting time and money.
  ...
  But I believe the biggest contributor to this trend is a misunderstanding of Karl Popper’s philosophy of science, which, to make a long story short, demands that a good scientific idea has to be falsifiable. Particle physicists seem to have misconstrued this to mean that any falsifiable idea is also good science.

• NYU Chemistry Professor Fired After Students Said His Class Was Too Hard

  According to The New York Times, 82 of Jones' 350 students signed the petition last spring; it alleged that too many of them were failing and that this was unacceptable. The students cited emotional and mental health complaints to make the case that Jones ought to make the class less difficult.
  ...
  The article does note that the petition never called for Jones to be fired. But the university evidently decided that the best way to resolve the situation was to turn him loose.

• Nose-picking primates spark scientific quest

  In a study they published in the Journal of Zoology, the team found 12 examples of primates caught in the nose-picking act.
  ...
  One study encouraged additional research by suggesting that the ingestion of nasal mucus could play an important role for the immune system, because of the immune proteins in the mucus.