recentpopularlog in

robertogreco : francisgalton   2

Physiognomy’s New Clothes – Blaise Aguera y Arcas – Medium
"In 1844, a laborer from a small town in southern Italy was put on trial for stealing “five ricottas, a hard cheese, two loaves of bread […] and two kid goats”. The laborer, Giuseppe Villella, was reportedly convicted of being a brigante (bandit), at a time when brigandage — banditry and state insurrection — was seen as endemic. Villella died in prison in Pavia, northern Italy, in 1864.

Villella’s death led to the birth of modern criminology. Nearby lived a scientist and surgeon named Cesare Lombroso, who believed that brigantes were a primitive type of people, prone to crime. Examining Villella’s remains, Lombroso found “evidence” confirming his belief: a depression on the occiput of the skull reminiscent of the skulls of “savages and apes”.

Using precise measurements, Lombroso recorded further physical traits he found indicative of derangement, including an “asymmetric face”. Criminals, Lombroso wrote, were “born criminals”. He held that criminality is inherited, and carries with it inherited physical characteristics that can be measured with instruments like calipers and craniographs [1]. This belief conveniently justified his a priori assumption that southern Italians were racially inferior to northern Italians.

The practice of using people’s outer appearance to infer inner character is called physiognomy. While today it is understood to be pseudoscience, the folk belief that there are inferior “types” of people, identifiable by their facial features and body measurements, has at various times been codified into country-wide law, providing a basis to acquire land, block immigration, justify slavery, and permit genocide. When put into practice, the pseudoscience of physiognomy becomes the pseudoscience of scientific racism.

Rapid developments in artificial intelligence and machine learning have enabled scientific racism to enter a new era, in which machine-learned models embed biases present in the human behavior used for model development. Whether intentional or not, this “laundering” of human prejudice through computer algorithms can make those biases appear to be justified objectively.

A recent case in point is Xiaolin Wu and Xi Zhang’s paper, “Automated Inference on Criminality Using Face Images”, submitted to arXiv (a popular online repository for physics and machine learning researchers) in November 2016. Wu and Zhang’s claim is that machine learning techniques can predict the likelihood that a person is a convicted criminal with nearly 90% accuracy using nothing but a driver’s license-style face photo. Although the paper was not peer-reviewed, its provocative findings generated a range of press coverage. [2]
Many of us in the research community found Wu and Zhang’s analysis deeply problematic, both ethically and scientifically. In one sense, it’s nothing new. However, the use of modern machine learning (which is both powerful and, to many, mysterious) can lend these old claims new credibility.

In an era of pervasive cameras and big data, machine-learned physiognomy can also be applied at unprecedented scale. Given society’s increasing reliance on machine learning for the automation of routine cognitive tasks, it is urgent that developers, critics, and users of artificial intelligence understand both the limits of the technology and the history of physiognomy, a set of practices and beliefs now being dressed in modern clothes. Hence, we are writing both in depth and for a wide audience: not only for researchers, engineers, journalists, and policymakers, but for anyone concerned about making sure AI technologies are a force for good.

We will begin by reviewing how the underlying machine learning technology works, then turn to a discussion of how machine learning can perpetuate human biases."



"Research shows that the photographer’s preconceptions and the context in which the photo is taken are as important as the faces themselves; different images of the same person can lead to widely different impressions. It is relatively easy to find a pair of images of two individuals matched with respect to age, race, and gender, such that one of them looks more trustworthy or more attractive, while in a different pair of images of the same people the other looks more trustworthy or more attractive."



"On a scientific level, machine learning can give us an unprecedented window into nature and human behavior, allowing us to introspect and systematically analyze patterns that used to be in the domain of intuition or folk wisdom. Seen through this lens, Wu and Zhang’s result is consistent with and extends a body of research that reveals some uncomfortable truths about how we tend to judge people.

On a practical level, machine learning technologies will increasingly become a part of all of our lives, and like many powerful tools they can and often will be used for good — including to make judgments based on data faster and fairer.

Machine learning can also be misused, often unintentionally. Such misuse tends to arise from an overly narrow focus on the technical problem, hence:

• Lack of insight into sources of bias in the training data;
• Lack of a careful review of existing research in the area, especially outside the field of machine learning;
• Not considering the various causal relationships that can produce a measured correlation;
• Not thinking through how the machine learning system might actually be used, and what societal effects that might have in practice.

Wu and Zhang’s paper illustrates all of the above traps. This is especially unfortunate given that the correlation they measure — assuming that it remains significant under more rigorous treatment — may actually be an important addition to the already significant body of research revealing pervasive bias in criminal judgment. Deep learning based on superficial features is decidedly not a tool that should be deployed to “accelerate” criminal justice; attempts to do so, like Faception’s, will instead perpetuate injustice."
blaiseaguerayarcas  physiognomy  2017  facerecognition  ai  artificialintelligence  machinelearning  racism  bias  xiaolinwu  xi  zhang  race  profiling  racialprofiling  giuseppevillella  cesarelombroso  pseudoscience  photography  chrononet  deeplearning  alexkrizhevsky  ilyasutskever  geoffreyhinton  gillevi  talhassner  alexnet  mugshots  objectivity  giambattistadellaporta  francisgalton  samuelnorton  josiahnott  georgegiddon  charlesdarwin  johnhoward  thomasclarkson  williamshakespeare  iscnewton  ernsthaeckel  scientificracism  jamesweidmann  faception  criminality  lawenforcement  faces  doothelange  mikeburton  trust  trustworthiness  stephenjaygould  philippafawcett  roberthughes  testosterone  gender  criminalclass  aggression  risk  riskassessment  judgement  brianholtz  shermanalexie  feedbackloops  identity  disability  ableism  disabilities 
may 2017 by robertogreco
Cyborgology: What is The History of The Quantified Self a History of?
[from Part 1: https://thesocietypages.org/cyborgology/2017/04/13/what-is-the-history-of-the-quantified-self-a-history-of-part-1/]

"In the past few months, I’ve posted about two works of long-form scholarship on the Quantified Self: Debora Lupton’s The Quantified Self and Gina Neff and Dawn Nufus’s Self-Tracking. Neff recently edited a volume of essays on QS (Quantified: Biosensing Technologies in Everyday Life, MIT 2016), but I’d like to take a not-so-brief break from reviewing books to address an issue that has been on my mind recently. Most texts that I read about the Quantified Self (be they traditional scholarship or more informal) refer to a meeting in 2007 at the house of Kevin Kelly for the official start to the QS movement. And while, yes, the name “Quantified Self” was coined by Kelly and his colleague Gary Wolf (the former founded Wired, the latter was an editor for the magazine), the practice of self-tracking obviously goes back much further than 10 years. Still, most historical references to the practice often point to Sanctorius of Padua, who, per an oft-cited study by consultant Melanie Swan, “studied energy expenditure in living systems by tracking his weight versus food intake and elimination for 30 years in the 16th century.” Neff and Nufus cite Benjamin Franklin’s practice of keeping a daily record of his time use. These anecdotal histories, however, don’t give us much in terms of understanding what a history of the Quantified Self is actually a history of.

Briefly, what I would like to prove over the course of a few posts is that at the heart of QS are statistics, anthropometrics, and psychometrics. I recognize that it’s not terribly controversial to suggest that these three technologies (I hesitate to call them “fields” here because of how widely they can be applied), all developed over the course of the nineteenth century, are critical to the way that QS works. Good thing, then, that there is a second half to my argument: as I touched upon briefly in my [shameless plug alert] Theorizing the Web talk last week, these three technologies were also critical to the proliferation of eugenics, that pseudoscientific attempt at strengthening the whole of the human race by breeding out or killing off those deemed deficient.

I don’t think it’s very hard to see an analogous relationship between QS and eugenics: both movements are predicated on anthropometrics and psychometrics, comparisons against norms, and the categorization and classification of human bodies as a result of the use of statistical technologies. But an analogy only gets us so far in seeking to build a history. I don’t think we can just jump from Francis Galton’s ramblings at the turn of one century to Kevin Kelly’s at the turn of the next. So what I’m going to attempt here is a sort of Foucauldian genealogy—from what was left of eugenics after its [rightful, though perhaps not as complete as one would hope] marginalization in the 1940s through to QS and the multi-billion dollar industry the movement has inspired.

I hope you’ll stick around for the full ride—it’s going to take a a number of weeks. For now, let’s start with a brief introduction to that bastion of Western exceptionalism: the eugenics movement."

[from Part 2: https://thesocietypages.org/cyborgology/2017/04/20/what-is-the-history-of-the-quantified-self-a-history-of-part-2/

"Here we begin to see an awkward situation in our quest to draw a line from Galton and hard-line eugenics (we will differentiate between hardline and “reform” eugenics further on) to the quantified self movement. Behaviorism sits diametrically opposed to eugenics for a number of reasons. Firstly, it does not distinguish between human and animal beings—certainly a tenet to which Galton and his like would object, understanding that humans are the superior species and a hierarchy of greatness existing within that species as well. Secondly, behaviorism accepts that outside, environmental influences will change the psychology of a subject. In 1971, Skinner argued that “An experimental analysis shifts the determination of behavior from autonomous man to the environment—an environment responsible both for the evolution of the species and for the repertoire acquired by each member” (214). This stands in direct conflict with the eugenical ideal that physical and psychological makeup is determined by heredity. Indeed, the eugenicist Robert Yerkes, otherwise close with Watson, wholly rejected the behaviorist’s views (Hergenhahn 400). Tracing the quantified-self’s behaviorist and self-experimental roots, then, leaves us without a very strong connection to the ideologies driving eugenics. Still, using Pearson as a hint, there may be a better path to follow."]

[from Part 3: https://thesocietypages.org/cyborgology/2017/04/27/what-is-the-history-of-the-quantified-self-a-history-of-part-3/

"The history of Galton and eugenics, then, can be traced into the history of personality tests. Once again, we come up against an awkward transition—this time from personality tests into the Quantified Self. Certainly, shades of Galtonian psychometrics show themselves to be present in QS technologies—that is, the treatment of statistical datasets for the purpose of correlation and prediction. Galton’s word association tests strongly influenced the MBTI, a test that, much like Quantified Self projects, seeks to help a subject make the right decisions in their life, though not through traditional Galtonian statistical tools. The MMPI and 16PFQ are for psychological evaluative purposes. And while some work has been done to suggest that “mental wellness” can be improved through self-tracking (see Kelley et al., Wolf 2009), much of the self-tracking ethos is based on factors that can be adjusted in order to see a correlative change in the subject (Wolf 2009). That is, by tracking my happiness on a daily basis against the amount of coffee I drink or the places I go, then I am acknowledging an environmental approach and declaring that my current psychological state is not set by my genealogy. A gap, then, between Galtonian personality tests and QS."]

[from Part 4 (Finale): https://thesocietypages.org/cyborgology/2017/05/08/what-is-the-history-of-the-quantified-self-a-history-of-the-finale/

"What is the history of the quantified self a history of? One could point to technological advances in circuitry miniaturization or in big data collection and processing. The proprietary and patented nature of the majority of QS devices precludes certain types of inquiry into their invention and proliferation. But it is not difficult to identify one of QS’s most critical underlying tenets: self-tracking for the purpose of self-improvement through the identification of behavioral and environmental variables critical to one’s physical and psychological makeup. Recognizing the importance of this premise to QS allows us to trace back through the scientific fields which have strongly influenced the QS movement—from both a consumer and product standpoint. Doing so, however, reveals a seeming incommensurability between an otherwise analogous pair: QS and eugenics. A eugenical emphasis on heredity sits in direct conflict to a self-tracker’s belief that a focus on environmental factors could change one’s life for the better—even while both are predicated on statistical analysis, both purport to improve the human stock, and both, as argued by Dale Carrico, make assertions towards what is a “normal” human.

A more complicated relationship between the two is revealed upon attempting this genealogical connection. What I have outlined over the past few weeks is, I hope, only the beginning of such a project. I chose not to produce a rhetorical analysis of the visual and textual language of efficiency in both movements—from that utilized by the likes of Frederick Taylor and his eugenicist protégés, the Gilbreths, to what Christina Cogdell calls “Biological Efficiency and Streamline Design” in her work, Eugenic Design, and into a deep trove of rhetoric around efficiency utilized by market-available QS device marketers. Nor did I aim to produce an exhaustive bibliographic lineage. I did, however, seek to use the strong sense of self-experimentation in QS to work backwards towards the presence of behaviorism in early-twentieth century eugenical rhetoric. Then, moving in the opposite direction, I tracked the proliferation of Galtonian psychometrics into mid-century personality test development and eventually into the risk-management goals of the neoliberal surveillance state. I hope that what I have argued will lead to a more in-depth investigation into each step along this homological relationship. In the grander scheme, I see this project as part of a critical interrogation into the Quantified Self. By throwing into sharp relief the linkages between eugenics and QS, I seek to encourage resistance to fetishizing the latter’s technologies and their output, as well as the potential for meaningful change via those technologies."]
gabischaffzin  quantifiedself  2017  kevinkelly  garywolf  eugenics  anthropometrics  psychometrics  measurement  statistics  heredity  francisgalton  charlesdarwin  adolphequetelet  normal  psychology  pernilsroll-hansen  michelfoucault  majianadesan  self-regulation  marginalization  anthropology  technology  data  personality  henryfairfieldosborn  moralbehaviorism  behaviorism  williamepstein  mitchelldean  neoliberalism  containment  risk  riskassessment  freedom  rehabilitation  responsibility  obligation  dalecarrico  fredericktaylor  christinacogdell  surveillance  nikolasrose  myers-briggs  mbti  katherinebriggs  isabelbriggsmeyers  bellcurve  emilkraepelin  charlesspearman  rymondcattell  personalitytests  allenneuringer  microsoft  self-experimentation  gamification  deborahlupton  johnwatson  robertyerkes  ginaneff  dawnnufus  self-tracking  melanieswan  benjaminfranklin  recordkeeping  foucault 
may 2017 by robertogreco

Copy this bookmark:





to read