Computerized Gaydar - a threat to privacy and safety of gay people

My Oct. 13, 2017 post dealt with controversy over an article referenced as forthcoming that showed neural networks to be more accurate than humans at detecting sexual orientation. It has now appeared, and I want to pass on its abstract here:

We show that faces contain much more information about sexual orientation than can be perceived or interpreted by the human brain. We used deep neural networks to extract features from 35,326 facial images. These features were entered into a logistic regression aimed at classifying sexual orientation. Given a single facial image, a classifier could correctly distinguish between gay and heterosexual men in 81% of cases, and in 71% of cases for women. Human judges achieved much lower accuracy: 61% for men and 54% for women. The accuracy of the algorithm increased to 91% and 83%, respectively, given five facial images per person. Facial features employed by the classifier included both fixed (e.g., nose shape) and transient facial features (e.g., grooming style). Consistent with the prenatal hormone theory of sexual orientation, gay men and women tended to have gender-atypical facial morphology, expression, and grooming styles. Prediction models aimed at gender alone allowed for detecting gay males with 57% accuracy and gay females with 58% accuracy. Those findings advance our understanding of the origins of sexual orientation and the limits of human perception. Additionally, given that companies and governments are increasingly using computer vision algorithms to detect people’s intimate traits, our findings expose a threat to the privacy and safety of gay men and women. 

Amygdala structure and defense of the social system.

Nam et al. examine the neuroanatomical substrates of preferences for maintaining existing social arrangements.

Individual variation in preferences to maintain versus change the societal status quo can manifest in the political realm by choosing leaders and policies that reinforce or undermine existing inequalities1. We sought to understand which individuals are likely to defend or challenge inequality in society by exploring the neuroanatomical substrates of system justification tendencies. In two independent neuroimaging studies, we observed that larger bilateral amygdala volume was positively correlated with the tendency to believe that the existing social order was legitimate and desirable. These results held for members of advantaged and disadvantaged groups (men and women, respectively). Furthermore, individuals with larger amygdala volume were less likely to participate in subsequent protest movements. We ruled out alternative explanations in terms of attitudinal extremity and political orientation per se. Exploratory whole-brain analyses suggested that system justification effects may extend to structures that are adjacent to the amygdala, including parts of the insula and the orbitofrontal cortex. These findings suggest that the amygdala may provide a neural substrate for maintaining the societal status quo, and opens avenues for further investigation into the association between system justification and other neuroanatomical regions.

Our microbiome challenges our concept of self.

Rees et al. offer an interesting perspective on the significance of our microbiome. Their abstract:

Today, the three classical biological explanations of the individual self––the immune system, the brain, the genome––are being challenged by the new field of microbiome research. Evidence shows that our resident microbes orchestrate the adaptive immune system, influence the brain, and contribute more gene functions than our own genome. The realization that humans are not individual, discrete entities but rather the outcome of ever-changing interactions with microorganisms has consequences beyond the biological disciplines. In particular, it calls into question the assumption that distinctive human traits set us apart from all other animals––and therefore also the traditional disciplinary divisions between the arts and the sciences.

Here from the discussion is their development (a bit shaky, I think) of the last point in the abstract:

Historically, the division of labor between faculties of arts and faculties of science emerged in the 18th century, alongside the idea that humans are more than mere nature––that there are human-exclusive capacities that set us apart from “mere” animals and plants. More specifically, the argument was that reason, language, and art had liberated the human from the contingencies of nature and had gradually given rise to a uniquely human world, a world of “culture” that is irreducible to the laws of nature and that therefore requires its own set of sciences (the term “culture” was first used to mark a distinctive human world in the late 1770s). Arguably, the findings of microbiome research profoundly trouble the comprehension of the human that has sustained the traditional distinction between the natural sciences (concerned with the nonhuman) and the arts (concerned with the human as more than mere nature). Provocatively put, if humans depend on microorganisms, then what is at stake in the study of microbes qua microbes is not only an understanding of microorganisms but also the human. This doesn’t mean that the field of the arts can now be conveniently ploughed in terms of the natural sciences. On the contrary, it means that the stakes of the natural sciences exceed the expertise of the natural sciences and reach over into the arts. This makes a close collaboration of the life sciences with the human sciences imperative.

As we see it, it is important but not enough to argue that we have never been individuals –– or to suggest that human and microbial worlds are inseparably entangled. What is needed, in addition, is a whole new configuration of research, one where arts and science are combined. The challenge is 2-fold. Researchers in the life sciences have to learn that the stakes of their research are bigger than their expertise, and researchers in the arts have to learn to think the human––philosophy, politics, and poetry––beyond the now untenable idea that humans are more than mere nature. The challenge is big, the opportunity even bigger: it is time, and perhaps past time, to rethink collaboratively––beyond arts and science divisions––what it means to be a living human being at home in a microbial world, one on which we depend and with which we are inseparably interwoven. Microbiome science has the exciting––the important––potential to catalyze the breakdown of the anachronistic barriers between the natural and the human sciences and enable a truly integrated understanding of what it means to be human, after the illusion of the bounded, individual self. The human is more than the human.

When our eyes move, our eardrums move.

Interesting stuff from Gruters et al.:

Significance

The peripheral hearing system contains several motor mechanisms that allow the brain to modify the auditory transduction process. Movements or tensioning of either the middle ear muscles or the outer hair cells modifies eardrum motion, producing sounds that can be detected by a microphone placed in the ear canal (e.g., as otoacoustic emissions). Here, we report a form of eardrum motion produced by the brain via these systems: oscillations synchronized with and covarying with the direction and amplitude of saccades. These observations suggest that a vision-related process modulates the first stage of hearing. In particular, these eye movement-related eardrum oscillations may help the brain connect sights and sounds despite changes in the spatial relationship between the eyes and the ears.

Abstract

Interactions between sensory pathways such as the visual and auditory systems are known to occur in the brain, but where they first occur is uncertain. Here, we show a multimodal interaction evident at the eardrum. Ear canal microphone measurements in humans (n = 19 ears in 16 subjects) and monkeys (n = 5 ears in three subjects) performing a saccadic eye movement task to visual targets indicated that the eardrum moves in conjunction with the eye movement. The eardrum motion was oscillatory and began as early as 10 ms before saccade onset in humans or with saccade onset in monkeys. These eardrum movements, which we dub eye movement-related eardrum oscillations (EMREOs), occurred in the absence of a sound stimulus. The amplitude and phase of the EMREOs depended on the direction and horizontal amplitude of the saccade. They lasted throughout the saccade and well into subsequent periods of steady fixation. We discuss the possibility that the mechanisms underlying EMREOs create eye movement-related binaural cues that may aid the brain in evaluating the relationship between visual and auditory stimulus locations as the eyes move.

Human cooperation in dynamic networks.

Melamed et al. show that the emergence of cooperative networks of humans does not require knowledge of the reputations of participants, only the severing of ties to non-cooperators.

Significance

Understanding the patterns and processes of human cooperation is of central scientific importance. Networks can promote cooperation when their existing or emergent topology allows conditional cooperators in the network to isolate themselves from exploitation by noncooperators. We do not know from prior work whether the emergent structures that promote cooperation are driven by reputation or can emerge purely via dynamics, i.e., the severing of ties to noncooperators and the formation of new ties irrespective of reputational information. Here we demonstrate, experimentally, that dynamic networks yield very high rates of cooperation even without reputational knowledge. Further, we identify realistic conditions under which static networks (where ties cannot be altered) yield cooperation rates as high as those in dynamic networks.

Abstract

Humans’ propensity to cooperate is driven by our embeddedness in social networks. A key mechanism through which networks promote cooperation is clustering. Within clusters, conditional cooperators are insulated from exploitation by noncooperators, allowing them to reap the benefits of cooperation. Dynamic networks, where ties can be shed and new ties formed, allow for the endogenous emergence of clusters of cooperators. Although past work suggests that either reputation processes or network dynamics can increase clustering and cooperation, existing work on network dynamics conflates reputations and dynamics. Here we report results from a large-scale experiment (total n = 2,675) that embedded participants in clustered or random networks that were static or dynamic, with varying levels of reputational information. Results show that initial network clustering predicts cooperation in static networks, but not in dynamic ones. Further, our experiment shows that while reputations are important for partner choice, cooperation levels are driven purely by dynamics. Supplemental conditions confirmed this lack of a reputation effect. Importantly, we find that when participants make individual choices to cooperate or defect with each partner, as opposed to a single decision that applies to all partners (as is standard in the literature on cooperation in networks), cooperation rates in static networks are as high as cooperation rates in dynamic networks. This finding highlights the importance of structured relations for sustained cooperation, and shows how giving experimental participants more realistic choices has important consequences for whether dynamic networks promote higher levels of cooperation than static networks.

Welcome to the post text future (of the new authoritarian state?)

White males (and the republican party) seem to perceive that the most effective way to maintain their current prominence as they continue to become a minority of the population is to enable movement towards an authoritarian state by suppressing minority voting and delegitimizing the institutions protecting democracy. Today’s Hungary provides an example of a single party ‘democracy’ led by a powerful leader. A fascinating sequence of articles in the New York Times describes additional powerful tools that have been moving public discourse from a rational textual context towards emotional visual video communication employing simple powerful memes. So far, the right has been more effective at exploiting such memes than the left. From Farhad Manjoo's article:

...An online culture ruled by pictures and sounds rather than text is going to alter much about how we understand the world around us.

The haze of misinformation hanging over online life will only darken under multimedia — think of your phone as a Hollywood-grade visual-effects studio that could be used to make anyone appear to say or do anything. The ability to search audio and video as easily as we search text means, effectively, the end of any private space.

Then there’s the more basic question of how pictures and sounds alter how we think. An information system dominated by pictures and sounds prizes emotion over rationality. It’s a world where slogans and memes have more sticking power than arguments. (Remind you of anyone?) And will someone please think of the children: Do you know how much power YouTube has over your kids? Are you afraid to find out?

But what are we going to do? There seems no going back now. For text, the writing is on the wall.

Bowles describes the mainstreaming of political memes online:

Groups like the conservative Look Ahead America and the liberal Center for Story-Based Strategy emerged to nurture memers, and big political donors like George Soros and the Mercer family funded meme efforts.

“It’s almost like a new means of communication — the image and emotion and creation,” said Matt Braynard, 39, the former director of data for Mr. Trump’s campaign, who is now the executive director of Look Ahead America. “I don’t want to call it literature, but it has an art.”

Organizers on both the left and right said the left has so far been slower to adapt to meme politics. To catch up, Sean Eldridge, husband of the Facebook co-founder Chris Hughes, is working on creating shareable content with Stand Up America, a progressive nonprofit that opposes President Trump. And the activist John Sellers’s The Other 98% has received funding from Open Society Foundations, a group backed by Mr. Soros.

Andrew Boyd, who designs campaigns for social change, was one of the first to document political memes, writing a seminal essay, “Truth is a Virus,” in 2002. He argued that the most important recent political meme has not come from either party’s campaign or donors but from the #MeToo movement around sexual harassment.

“It has a crystal quality to it, a simplicity, and elegance, something that feels right and organized,” said Mr. Boyd, 55. “Me too. Me too. That happened to me too. The best memes are very populist, and yet they have a precision.”

I won't pass on any further chunks, but strongly urge MindBlog readers to check out the sequence of articles in the NY Times special segment.

Ancient origins of walking on land - the walking skate.

From Jung et al., in the Feb. 8 issue of Cell, report finding that the circuits underlying vertebrate ambulation on dry land appear in a skate many millions of years before fish crawled out of the prehistoric ocean.




Highlights

•The little skate Leucoraja erinacea exhibits bipedal walking-like behaviors 

•Neuronal subtypes essential for walking originated in primitive jawed fish 

•Fin and limb motor neurons share a common Hox-dependent gene network 

•Modulation of Hox patterning facilitates evolutionary changes in MN organization

Summary

Walking is the predominant locomotor behavior expressed by land-dwelling vertebrates, but it is unknown when the neural circuits that are essential for limb control first appeared. Certain fish species display walking-like behaviors, raising the possibility that the underlying circuitry originated in primitive marine vertebrates. We show that the neural substrates of bipedalism are present in the little skate Leucoraja erinacea, whose common ancestor with tetrapods existed ∼420 million years ago. Leucoraja exhibits core features of tetrapod locomotor gaits, including left-right alternation and reciprocal extension-flexion of the pelvic fins. Leucoraja also deploys a remarkably conserved Hox transcription factor-dependent program that is essential for selective innervation of fin/limb muscle. This network encodes peripheral connectivity modules that are distinct from those used in axial muscle-based swimming and has apparently been diminished in most modern fish. These findings indicate that the circuits that are essential for walking evolved through adaptation of a genetic regulatory network shared by all vertebrates with paired appendages.

Computers no better than humans at predicting who should go to jail.

I want to point to two pieces that offer a commentary on our trust in using algorithms to predict human behavior rather than old fashioned human judgement. In the U.S., computers help decide who goes to jail, on the basis of predicting recidivism. Matacic discusses studies showing their judgment may be no better than ours. She points, for example, to work of Dressel and Farid. Their abstract:

Algorithms for predicting recidivism are commonly used to assess a criminal defendant’s likelihood of committing a crime. These predictions are used in pretrial, parole, and sentencing decisions. Proponents of these systems argue that big data and advanced machine learning make these analyses more accurate and less biased than humans. We show, however, that the widely used commercial risk assessment software COMPAS is no more accurate or fair than predictions made by people with little or no criminal justice expertise. We further show that a simple linear predictor provided with only two features is nearly equivalent to COMPAS with its 137 features.

Afraid of A.I.? Be very afraid of quantum computing....

Several recent articles point to a quiet revolution that is receiving relatively little notice - the development of quantum computers that may operate 100,000 times faster than today's binary computers that store information as 0's and 1's, by using quantum entanglement and superposition to process information. A description from Wadhwa's article:

Unlike classic computers, in which information is represented in 0’s and 1’s, quantum computers rely on particles called quantum bits, or qubits. These can hold a value of 0 or 1 or both values at the same time — a superposition denoted as “0+1.” They solve problems by laying out all of the possibilities simultaneously and measuring the results. It’s equivalent to opening a combination lock by trying every possible number and sequence simultaneously.

Friedman quotes IBM researcher Talia Gershon, who

...posted a fun video explaining the power of quantum computers to optimize and model problems with an exponential number of variables. She displayed a picture of a table at her wedding set for 10 guests, and posed this question: How many different ways can you seat 10 people? It turns out, she explained, there are “3.6 million ways to arrange 10 people for dinner.”

Classical computers don’t solve “big versions of this problem very well at all,” she said, like trying to crack sophisticated encrypted codes, where you need to try a massive number of variables, or modeling molecules where you need to account for an exponential number of interactions. Quantum computers, with their exponential processing power, will be able to crack most encryption without breaking a sweat.

From Wadhwa:

IBM is already offering early versions of quantum computing as a cloud service to select clients. There is a global race between technology companies, defense contractors, universities and governments to build advanced versions that hold the promise of solving some of the greatest mysteries of the universe — and enable the cracking open of practically every secured database in the world.

Artificial intelligence applications based on conventional computing and already in the pipeline are going to bring about a world-wide decimation of jobs in the next 10-20 years. What will the effect of exponentially more powerful computers be? IBM recently upgraded its publicly available processor to 20 qubits and has the operational prototype of a 50-qubit processor. A 50-qubit computer would exceed the capability of the top supercomputers in the world, and a 100-qubit computer is a possibility.

A Valentine's Day musical offering - Respighi's Intermezzo Serenata

A piano transcription by Respighi of the Intermezzo Serenata from his comic opera Re Enzo.  Making a recording of a piano piece I like motivates me to learn it a bit more thoroughly and is necessary if I plan to play it for one of our house social/musical occasions.  I never manage a playing without some minor error,  so from the four takes I did yesterday I pass on with one with the smallest number of glitches. Readers who are Apple groupies might note the Apple watch camera app displaying the scene being recorded by the iPhone X camera and the shutter button on the watch starting and stopping the recording.

Mind treats on the web.

This MindBlog is a miniscule drop in a large internet pond of excellent and interesting sites describing new mind research and ideas. I am aware of only a fraction what is out there. This brief post is to pass on a few random mentions. A recent email asks me to pitch a new drop in the pond, Neo.life, that describes itself as an online magazine covering experiments we’re doing on ourselves, tracking ideas and technologies that enable us to analyze, enhance, and edit ourselves (see, for example “When algorithms are running the asylum").” Aeon.co offers a daily email with occasional articles that I find so interesting that I have sent them a modest contribution. Edge.org is a very rich site whose annual question has provoked many thoughtful essays, although I think this year’s annual question (‘What is the last question?’) is a loser. And, of course there are the well known TED talks. When I find a talk on mind matters that I think might be interesting I go straight to the text transcript, not being patient enough to watch the video.

My apologies to many other sites I am not mentioning.

“The rich are different from you and me..” - the psychology of inequality

In the mythical exchange between F. Scott Fitzgerald and Ernest Hemingway, Hemingway responds “Yes, they have more money.” Who are ‘they’? … presumably those who can be unconcerned about the stock market ‘correction’ over the past week. (Cohen brings some perspective the market drop by noting that 84% of all stocks owned by Americans - including all retirement and savings assets - belong to the wealthiest 10 percent of households, for whom a market drop of 10% would bring only a 1-2% change in their wealth holdings.)

In this vein, I would like to point to Elizabeth Kolbert's fascinating essay on the differences between being on the lower rather than the upper part of the economic ladder. She cites psychological studies showing that much of the damage done by being poor comes from feeling poor (as, for example when people find they are making less than their peers), not from actual material deprivations. "In a world where people measure themselves against their neighbors, it possible to earn good money and still feel deprived...thanks to the growing gap between the one percent and everyone else, the subjective effect is of widespread impoverishment...inequality mimics poverty in our minds.

Studies show that feeling disadvantaged leads people to do more risk taking, amplify perception of racial differences, be more susceptible to conspiracy theories, and be more likely to have medical problems.

If feeling poor goes with feeling bad about it, it would seem reasonable to suppose that rich people feel good about being rich. Work by Rachel Sherman shows that this is not the case. The few rich people who were willing to be interviewed by her didn't want their wealth to be known and preferred not to think of themselves as privileged, citing friends who even more wealthy. "If affluence is in the eye of the beholder, then even the super-rich, when they compare their situation with that of the ultra-rich, can feel sorry for themselves." Also, they also feel moral conflicts about having privilege in general.

Preschoolers, brown capuchin monkeys, California state workers, college students recruited for psychological experiments—everyone, it seems, resents inequity. This is true even though what counts as being disadvantaged varies from place to place and from year to year...Thomas Jefferson, living at Monticello without hot water or overhead lighting, would, by the standards of contemporary America, be considered “poorer than the poor.” No doubt inequity, which, by many accounts, is a precondition for civilization, has been a driving force behind the kinds of innovations that have made indoor plumbing and electricity, not to mention refrigeration, central heating, and Wi-Fi, come, in the intervening centuries, to seem necessities in the U.S.

Still, there are choices to be made. The tax bill recently approved by Congress directs, in ways both big and small, even more gains to the country’s plutocrats. Supporters insist that the measure will generate so much prosperity that the poor and the middle class will also end up benefitting. But even if this proves true—and all evidence suggests that it will not—the measure doesn’t address the real problem. It’s not greater wealth but greater equity that will make us all feel richer. 

Our body tissues crosstalk during exercise.

Gretchen Reynolds points to fascinating work showing that exercise causes cells to release tiny lipid coated hollow spheres (vesicles) containing protein molecules that carry signals between muscle cells, fat cells, and the liver cells that generate energy during exercise - regulating coordination between organs during exercise. 

Highlights

•Exosomes and small vesicles are released into circulation with exercise 

•Proteins without a signal peptide sequence circulate in vesicles during exercise 

•Exercise-liberated vesicles have a propensity to localize in the liver 

•Femoral arteriovenous difference analysis identifies 35 novel candidate myokines

Summary

Exercise stimulates the release of molecules into the circulation, supporting the concept that inter-tissue signaling proteins are important mediators of adaptations to exercise. Recognizing that many circulating proteins are packaged in extracellular vesicles (EVs), we employed quantitative proteomic techniques to characterize the exercise-induced secretion of EV-contained proteins. Following a 1-hr bout of cycling exercise in healthy humans, we observed an increase in the circulation of over 300 proteins, with a notable enrichment of several classes of proteins that compose exosomes and small vesicles. Pulse-chase and intravital imaging experiments suggested EVs liberated by exercise have a propensity to localize in the liver and can transfer their protein cargo. Moreover, by employing arteriovenous balance studies across the contracting human limb, we identified several novel candidate myokines, released into circulation independently of classical secretion. These data identify a new paradigm by which tissue crosstalk during exercise can exert systemic biological effects.

A Thursday musical offering... Charles Meyer - Valse Melancolique

Under the 'random curious stuff' in MindBlog's subtitle, above, I've decided to resume recording some of the pieces that I enjoy playing, using only my iPhone X with a small condenser microphone plugged in to its USB port.  This lyrical piece is frequently incorrectly attributed to Chopin.