automating ineq

automating ineq.png

by Virginia Eubanks

suggested to library to purchase.. and on hold.. in the mean time..  (thank you library – notes at bottom of page) keeps coming up to adding page ahead of reading it..

_________

more notes on Virginia’s page.. but adding this bit from data & society talk here

j: diff is.. judge could be biased.. but systematizing bias is concrete

today we have the means to have no one play judge..let’s do that

(tech) as it could be2 convos .. as the day

ie: hlb via 2 convos that io dance.. as the day..[aka: not part\ial.. for (blank)’s sake…]..  a nother way

q: can any of us live under that surveillance.. i think i could live in a world where we all knew everything about everyone.. but it’s not symmetric like that..

gershenfeld sel

v: if we build a world w contempt/hatred for poor people .. that’s going to affect all of us

interconnectedness law

_________

Lisa Cunningham (@LisaGemini) tweeted at 7:49 PM on Tue, Jan 23, 2018:
Not homeless enough? What does that mean? This is why I hate bureaucrats. #Homeless children are the ones who suffer! https://t.co/ODYf0aT2SO
(https://twitter.com/LisaGemini/status/955996021653467138?ref_src=twcamp%5Eshare%7Ctwsrc%5Eandroid%7Ctwgr%5Egmail%7Ctwcon%5E7090%7Ctwterm%5E0)

________

notes/quotes from book:

intro

3

health care fraud is a real problem. according to the fbi, it costs employers, policy holders, and taxpayers nearly 30 bn a year. though the great majority of it is committed by providers, not consumers..  i don’t fault ins co’s for using the tools at their disposal to id fraudulent claims, or even for trying to predict them. but the human impacts of red flagging, esp when it leads to he loss of crucial life saving services.. can be catastrophic.

insuranceness

5

the ins co repeatedly told me the problem was the result of a tech error, a few missing digits in a database.. but that’s the thing about being targeted by an algo.. you get a sense of a pattern in the digital noise, and electronic eye turned toward you, but you can’t put your finger on exactly what’s amiss.. there is not req that you be notified when you are red flagged.. (she notes legit suspicions of their total legit situation)

we all inhabit this new regime of digital data, but we don’t all experience it in the same way. what made my fam’s experience endurable was the access to info.. discretionary time, and self determination that professional middle class people often take for granted.. i knew enough about algo decision making to immediately suspect that we had been targeted for a fraud investigation..  we never stopped assuming we were eligible for med ins.. so jason got the care he needed..

6

perhaps most importantly, the kind of digital scrutiny jason and i underwent is a daily occurrence for many people, not a one time aberration..

7

the data acts to reinforce their marginality when it is used to target them for suspicion and extra scrutiny.. those groups seen as undeserving are singled out for punitive public policy and more intense surveillance, and the cycle begins again.. it is a kind of collective red flagging, a feedback loop of injustice..

8

this is performative politics. the legislation was not intended to work; it was intended to heap stigma on social programs and reinforce the cultural narrative that those who access public assistance are criminal, lazy, spendthrift addicts.. (giving out money in maine 2014 – then tracking – not how it was spent – but where it was withdrawn)

lepage’s use of ebt data .. to track/stigmatize poor and working class people’s decision making didn’t come as much of a surprise to me. by 2014.. i had been thinking and writing about tech and poverty for 20 yrs..

9

techs of poverty management are not neutral. they are shaped by our nation’s fear of econ insecurity and hatred of the poor.. they in turn shape the politics and experience of poverty

poor women are the test subjects for surveillance tech dorothy told me.. you should pay attention to what happens to us.. you’re next..

10

so in 2014.. i set out to investigate .. i chose 3 stories to explore: 1\ indian’s welfare system; 2\ la unhoused electronic registry; 3\ pennsylvania model to predict potential children of abuse/neglect

11

what i found was stunning.. life threatening consequences.. discourage from claiming resources.. collect personal info while offering nothing much in return.. predictive algos tag them as risky w problematic parents..  every move visible..

12

automated eligibility is now standard practice in almost every state’s public assistance office.. et al..  the widespread use of these systems impacts the quality of democracy for us all..

automated decision making shatters the social safety net, criminalizes the poor, intensifies discrimination and compromises our deepest national values

sounds like voting..

begs we redefine decision making.. disengage from consensus ness
ie: hosting life bits within short ness.. via  2 convos that io dance.. as the day..
systems first designed for the poor will eventually be used on everyone.. digital poorhouse.. part of a long american tradition. we manage the individual poor in order to escape our shared responsibility for eradicating poverty

1 – from poorhouse to database

16

while poorhouses have been physically demolished, their legacy remains alive and well in the automated decision making systems that encage and entrap today’s poor..

17

first poorhouse.. 1662..  not until 1820s that imprisoning the indigent in public institutions became the nations’ primary method of regulating poverty..

19

by 1856 about 1/4 of poorhouse residents in ny were children. another 2/4 were mentally ill, blind, deaf, or developmentally delayed.. most of the rest were elderly, ill, physically disabled or poor mothers recovering from childbirth

22

caseworkers assume that the poor were not reliable witnesses. they confirmed their stories w police, neighbors, local shopkeepers, clergy, schoolteachers, nurses.. scientific charity treated the poor as criminal defendants by default..

sci charity workers advised in depth investigation of apps for relief because they believe that there was a hereditary division between deserving and undeserving poor whites..

24

the scientific charity movement relied on a slew of new inventions: the caseworker, the relief investigation, the eugenics record, the data clearinghouse..  it drew on what lawyers, academics, and drs believe to be the most empirically sophisticated science of its time.. but were systems of disempowering poor and working class people.. if the poorhouse was a machine that diverted the poor and working class from public resources, scientific charity was a technique of producing plausible deniability in elites..t

wow.. a bunch of pages are out of order.. fine to 20 then.. ie: 21, 24, 23, 22, 27, 26, 25, 28, 29, 32, 31, 30, 35, 34, 33, 36, 37, 40, 39, 38, 43, 42, 41, 44, 45, 48, 47, 46,, 51, 50, 49, 52, 53, 54, 55, the rest fine

28

by abandoning the idea of a universal benefits program, roosevelt resurrected sci charity’s investigation, policing and diversion.. but rather than being directed at a broad spectrum of the poor and working class, these techniques were selectively applied to a new target group what was just emerging. they would come to be known as ‘welfare mothers’

29

welfare required that poor people trade their rights – to bodily integrity, safe work environments, mobility, political participation, privacy and self-determination – for meager aid for their families..

discriminatory eligibility rules gave caseworker broad latitude .. ie: 63 in alameda cty ca.. invade homes of 700 welfare recipients..  failed to id selves.. broke doors.. conducted mainly against poc

30

60s.. most importantly members of the movement insisted that motherwork is work..

interpretive labor.. and work

31

afdc became so embattled that nixon proposed a guaranteed annual income program..  1600 a year for  family of four.. to families excluded from afdc..  it also included built in work requirements.. this was  a sticking point for single mothers w small children..  it failed.. and pressure on afdc continued to mount..

32

jonnie tillmon, first chairwoman of the nwro recognized that white welfare recipients were fellow sufferers and potential allies.. as she explained in a 1971 interview, ‘we can’t afford racial separateness. i’m told by the poor white girls on welfare how they feel when they’re hungry ,and i feel the same way when i’m hungry’

but afdc drummed up white middle class animosity to turn back the movement’s success..

33

driven by reagan and other conservative politicians, a taxpayer revolt against afdc challenged the notion that the poor should have full complement of rights promised by the constitution.. but the welfare rights movement’s successes were enshrined into law, so exclusion from public assistance could not longer be accomplished thru discriminatory eligibility rules..

elected officials and state bureaucrats, caught between increasingly stringent legal protections and demands to contain public assistance spending, performed a political sleight of hand.. they commissioned expansive new techs that promised to save money by distributing aid more *efficiently.. in fact, these tech systems acted like walls, standing between poor people and their legal rights.. in this moment, the digital poorhouse was born.. shrinking spending by increasing scrutiny/surveil

efficiency ness

36

(rockefeller helps institute/enforce this auto ineq)

in 1973 nearly half of people living under poverty line in us received afdc. a decade later, after new techs.. dropped to 30 %.. today it is less than 10%..t

37

on claims new gen of digi tools as ‘disruptive’ .. but when focus on programs specifically targeted at poor and working clas.s. new regime of data anal is more evolution than revolution.. simply an expansion and continuation of moralistic and punitive poverty management strategies that have been w us since the 1820s

38

poor and working class resist restriction of their rights, dismantle discriminatory institutions and join together for survival and mutual aid. but time and again they face middle class backlash..  social assistance is recast as charity, mutual aid is reconstructed as dependency,

a well funded, widely supported and wildly successful countermovement to deny basic human rights to poor and working class has grown steadily since the 70s.. the movement manufactures and circulates misleading stories about the poor: undeserving, fraudulent, dependent, immoral..  very effective propaganda.. to convince americans that working class and poor must battle each other in a zero sum game over limited resources.. more quietly, program admin and data scientists push high tech tools that promise to help more.. more humanely, while promoting efficiency.. id ing fraud and containing costs..  framed as way to rationalize and streamline benefits.. but real goal is what it has always been: to profile, police, and punish the poor

manufacturing consent perpetuating poverty

2 – automating eligibility in the heartland

51

between 2006 and 2008 the state of indiana denied more than a million apps for food stamps, medicaid and cash benefits, a 54% increase compared to the three yrs prior to automation

benefits discontinued for ‘failure to coop in establishing eligibility’

52

the old system involved caseworkers developing one on one relationships.. and following cases thru to completion.. the new system presented call center workers w a lists of tasks rather than a docket of families.. no one worker had oversight of a case from beginning to end.. always spoke to a new worker..

62

from caseworker: you were expected to produce, and you couldn’t do that if you listened to the client’s story

63

another caseworker: the rule became brittle.. if applicants didn’t sense something in, one of thirty docs.. you simply closed the case for failure to comply.. you couldn’t go out of your way to help somebody

78

suitable home.. employable mother.. to block african american women from claiming benes …man in house.. sub father.. legitimized intrusion into their privacy/homes.. judgment of their sexuality..

82

when governor signed contract w ibm in 2006.. 38% of poor families w children were receiving cash benes from tanf. by 2014. the number had dropped to 8%

83

in the end, the indiana automation experiment was a form of digital diversion for poor and working americans..

3 – high tech homelessness in the city of angels

90

since 1950, more than 13 000 units of low income housing have been removed from skid row, enough for them all (homeless in la).. t

home less ness

iwan baan ness

94

(on getting people in homes faster.. and taking highly personal info) if survey takers request the more complete privacy notice, they learn that their info will be shared w 168 diff orgs.. including lapd.. the consent is valid for 7 yrs

101

no one told gary he would have to have a  3-5 yr verifiable rental history and a good credit history in order to qualify for their waiting list.. ‘how is that all relevant to getting housing for the homeless?’

105

according to the 2017 homeless count for greater la, 75% of the homeless in s la are completely unsheltered..  while 2364 unhoused people find shelter beds or permanent supportive hosing, another 6879 live in the makeshift shelters that have become s la’s de facto source of low income housing.. 70% of them are black

106

pathways is officially a 90 day center but getting people housed in 3 months is nearly impossible..

107

on guy scoring 1 out of 17 .. then reassessed and scored 16

the housing authority can be very very tricky.. if pathways client scores a 16.. he should qualify for shelter plus care voucher providing both rental assistance and supportive social services.. but then the housing authority says… you’re not really capable of living independently.. go and get something from a dr or psych letting us know you won’t put some water to boil and burn down the building..   it seems the housing authority wants to interview you out of services… he (menjivar) says..  whereas i’ve interviewed you into services (on pathways listening to stories.. but system still gets you).. so we tell them.. pretend you’re on trial… don’t divulge any additional info

108

if get section 8 vouchers.. only last 6 months.. and can take longer than that to find a place.. where landlord will take them in.. so process starts all over..

109

so.. they got all this info to create this database, talk about how many thousands of people are homeless, but never come back to serve them – richard renteria

in the absence of sufficient public investment in building or repurposing housing, coordinate entry is a system for managing homelessness, not solving it.. t

w/o increasing resources.. we don’ solve homelessness.. molly rysman.. homes/housing deputy for la county.. coordinated entry has made us much more efficient. but there’s no chance of ending homelessness w/o resources..

112

ko pointed out that coordinated entry allowed members of the ces network to arrive at city council  and board w … numbers.. but the real driver .. was the spread of tent cities..

according to lahsa’s 2017 homeless count.. there are 57 794 unhoused people in la county…. since 2014, managed to survey 31 124.. 35-50%.. of those.. managed to connect 9627 w housing.. ko estimates that coord entry has cost approx 11 mn.. not including cost of providing actual housing/services..  ces eased the way to some kind of housing resource for 17% at a cost of approx 1140 per person.. it is easy to argue that this is money well spent.

?

114

entry into the system.. w no where for them to go

some suspect that all that data is being held for other purposes entirely: to surveil and criminalize the unhoused. as of this writing.. the protected personal info of 21 500 of la’s most vulnerable people remains in a database that may never connect them w life saving services.. t.. it is possible to revoke your consent .. but the process is complicated.. even after expungement, some data stays in the system..

115

in theory it is still possible to access resources while refusing to supply protected personal info but the united way concedes that they are ‘not sure how many people use this option’

116

on criminalizing poor in america..the most direct parallel is operation talon, a joint effort of the office of inspector general and local welfare offices.. that mined food stamp data to id those w outstanding warrants (ie: from tickets not paid for sleeping outside et al), and then lured them to appointments regarding their benefits. when targeted recipients arrived at the welfare office, they were arrested..

this kind of blanket access to deeply personal info makes little sense outside of a system that equates poverty and homelessness w criminality..t

117

before jones and lavan injunctions, skid row was one of most policed neighborhood in world.. safer city initiative.. earmarked 6 mn annually to target status crimes associated w homelessness: sitting on sidewalk, jaywalking, littering, camping, and panhandling

118

according to urban sociologist forrest stuart, lapd officers made roughly 9 000 arrests and issued 12 000 citation in the first year of the initiative..  in an area w only 12 000 to 15 000 residents..

119

the constant data collection.. create what skid row residents perceive as a net of constraint that influences their every decision. daily, they feel encouraged to self-deport or self-imprison..t

120

charges for crimes associated w homelessness are often dismissed when cases come to trial, but in the meantime skid row residents might spend three or four months locked up. as a result, they lose their housing, their documents, their few possessions, and are passed over for social services.

122

in contrast (to targeting a group to surveil ie: cointelpro) in new data based surveil.. the target often emerges from the data

123

if homelessness is inevitable – like a disease or a natural disaster – then it is perfectly reasonable to use triage oriented solution s that prioritize unhoused people for a chance at limited housing resources. but if homelessness is a human tragedy created by policy decision and professional middle class apathy, coord entry allows us to distance our selves from the human impacts of our choice to not act decisively….

124

the problem is not that the city lacks adequate data on what kind of housing is needed to address the homelessness problem. rather, poor and working class people and their allies may not be able to overcome explicit political resistance from organized elites..

the proponents of the coord entry system, like many who seek to harness computational power for social justice, tend to find affinity w systems engineering approaches to social problems.. in this model.. political conflict arises primarily from lack of info..

125

having more info won’t necessarily resolve them…

blasi: homelessness is not a systems engineering problem. it’s a carpentry problem

126

likely his score will lower.. the model counts prison as housing..  he’ll stay trapped, too vigorous for intervention and too marginal to make a go of it w/o support. gary: ‘i’m  a criminal just for existing on the face of the earth’

4 – the allegheny algorithm

130

3/4 of child welfare investigation involve neglect rather than physical, sexual, or emotional abuse.. unhoused families face particularly difficult challenges holding on to their children, as the very condition of being homeless is judged neglectful

134

when i cam to run children and youth, it was a national disgrace, said marc cherna.. when he arrived in 1996 there were 1600 children waiting to be adopted, and the agency was only managing to process 60 adoptions a year..  70% of children .. in foster care system were black, those african americans made up only 11% of population of allegheny county..

137

(on cherna building community trust and turning things around.. then huge funding cut.. so they sought tool to do triage decision making).. tool that would mine cherna’s warehoused data to make prediction about which allegheny county children might be at greatest risk for abuse/neglect..

a predictive model using 132 variables..

138

when new zealand public learned of the project in 2014.. they responded w concern. academic researcher warned that the model might not be as accurate as team claimed: it was wrong about nearly 70% of the children it id’d as at highest risk of harm in the historical data.. others cautioned that model was primarily a tool of surveillance of the poor.. the experiment collapses in face of public resistance. but by that time, the vaithianathan team had won the contract to create a similar predictive risk model in allegheny county..

142

the afst is supposed to support, not supplant, human decision making in the call center. and yet, in practice, the algo seems to be training the intake workers..

a 14 yr old living in a cold and dirty house gets a risk score almost three times as high as a 6 yr old whose mother suspected he may have been abused and who may not be homeless..  in these cases, the model doesn’t seem to meet a commonsense standard .. why..?

cathy o neil has written that ‘models are opinions embedded in math’

cathy

models are useful. let us strip extraneous info and focus on what’s most critical to outcomes.. but they are also abstractions.. .. reflect priorities/preoccupation of their creators.. human decision making is reflected in thee key components of the afst: outcome variables, predictive variables, and validation data..

144

so the afst actually predicts decision made by the community (which families will be reported to the hotline) and by the agency and the family courts (which children will be removed from their families), not which children will be harmed..

145

in 2016, there were 15139 reports of abuse and neglect in allegheny cty. at its current rate of accuracy, the afst would have produced 3633 incorrect predictions

146

a model’s predictive ability is compromised when outcome variable are subjective..

152

ie: in 2011 51% of children in foster care in alaska were native american.. though native americans only 17% of youth population.. in illinois.. 53% of children in foster care were african american… though african americans make up only 16% of youth pop

153

in 2016.. 48% of children in foster care in allegheny county were african american, thought they made up only 18% of county’s children and youth..

a 2010 study in allegheny cty found that the great majority of disproportionality .. arises from referral bias, not screening bias..

154

the afst focuses all its predictive power and computation might on call screening, the step it can experimentally control, rather than concentrating on referral.. the step where racial disproportionality is actually entering the system..

155

of the 15139 reports.. in 2016.. can estimate that 605 were intentionally false.. is it illegal to call a malicious reports.. but pennsylvania accepts reports from anonymous callers.. so little a parent can do

the activity that introduces the most racial bias into the system is the very way the model defines maltreatment.. this easily gameable, discriminatory variable threatens to reverse all of the extraordinary work cherna and his team have done..

though only 27% of pittsburgh children receive public assistance, 80% of children placed in foster care in 2015 were removed from families relying on temp assistance for needy families.. (tanf) or supplemental nutrition assistance program (snap)..  that is, in allegheny county, class based disproportionality is worse than racial disporportionality.. but unlike other historically disadvantaged groups, the poor are not widely recognized as a legally protected class, so the disproportionate and discriminatory attention paid to poor families by child welfare offices goes largely unchallenged..t

the afst sees the use of public services as a risk to children..

the overwhelming majority of child welfare investigation in the us involve neglect, not abuse..t .. ie: of the 3.4 mn children involved in child welfare investigation in 2015.. 75% were for neglect.. while only 1/4 were for physical, emotional or sexual abuse..

157

nearly all the indicators of child neglect are also indicators of poverty:..t.. lack of food, inadequate housing, unlicensed childcare, unreliable transportation, utility shutoffs, homelessness, lack of health care..

child welfare services are not means tested: you don’t have to be low income to access them.. but professional middle class families rely instead on private sources for family support, so their interactions w helping professional are not tracked or represented in the data warehouse..

it is interesting to imagine the response if allegheny cty proposed including data from nannies, babysitters, private therapists, aa, and luxury rehabilitation centers to predict child abuse among wealthier families.. .. the professional middle class will not stand for such intrusive data gathering..

families avoid cyf if they can afford to, because the agency mixes two distinct and contradictory roles: provider of family support and investigator of maltreatment..

158

accepting resources means accepting the agency’s authority to remove your children. this is an invasive, terrifying trade off that parents w other option are not likely to choose.. poor and working class families feel forced to trade their rights to privacy, protection from unreasonable searches and due process for a chance at the resources and services they need to keep their children safe..

poverty is incontrovertibly harmful to children.. it is also harmful to their parents..

because the model confuses parenting while poor w poor parenting.. t.. the afst views parents who reach out to public programs as risks to their children

dang

159

eventually, the agency required her to give up her son to access the basic material resources that would have allowed her to care for him effectively herself..t

162

parenting while poor means parenting in public.. the state of penn’s goal for child safety, ‘being free from immediate physical or emotional harm.. can be difficult to reach, even for well resourced families..

163

a report of abuse or neglect that is found credible has profound impact on parent’s life for decades.. most jobs won’t hire.. w/in 90 days can request an admin review to amend or expunge record..  not many (poor families) dare to take cyf on in court..  attorney and pro bono program coordinator .. told me she’s neer rep’d a client in a cyf expungement case..  in part because it is a ‘much more extensive process than criminal expungement’ .. and if it does happen (expunged).. parent remains in state abuse registry until child turns 23

164

the expunge process applies only to those reported for grievous neglect or abuse.. any allegation that involve ‘non serious injury or neglect’ are referred to gen protective services.. gps data is kept indefinitely.. so the multiple calls on harriette, angel’s feisty but mostly obedient daughter? there is no way to expunge them.. eve tough they were clearly nuisance calls.. when and if harriette becomes a mom, she’ll start out w a higher afst score, because she had interactions w the child protective system as a kid..

166

prof middle class reach out for support all the time.. but because it is all privately funded.. not in data warehouse.. the same willingness to reach out for support by poor and working class.. because they are asking for public resources, labels them a risk to their children in the afst, even though cyf sees requesting resources as a positive attribute of parents..

cyf involved families.. would rather have an imperfect person making decisions about their families tha a flawless computer.. ‘you can teach people how you want to be treated.. opp to fix it w a person.. you can’t fix that number’

168

(on using algo tools on claims that there’s no way to know what’s going on in heads of caseworkers.. homeless provider..for decision making).. the philosophy that sees human beings as unknowable black boxes and machines as transparent deeply troubling. it seems to me a worldview that surrenders any attempt at empathy and forecloses the possibility of ethical development.. the presumption that human decision making is opaque and inaccessible is and admission that we have abandoned a social commitment to try to understand each other

let’s go deeper.. abandoned a social commitment to our interconnectedness.. our commons.. ness.. (just saying.. people or algo’d-tool.. aren’t our only two choices.. we can decide to let go of control and get to the root here)

169

according to us centers for disease control’s division of violence prevention, the largest risk factors for the perpetration of child abuse and neglect include social isolation, material deprivation, and parenting stress, all of which increase when parents feel watched all the time.. lose resources they need, suffer stigma, or are afraid to reach out to public programs for help..  a horrible irony is that the afst might create the very abuse it seeks to prevent

no doubt

equity – everyone getting a go everyday

as it could be

ie: hlb via 2 convos that io dance.. as the day..[aka: not part\ial.. for (blank)’s sake…]..  a nother way

it is difficult to say a predictive model works if it produces the outcome it is trying to measure.. thanks in part to the higher risk score, the parent is targeted for more punitive treatments.. must fulfill more agency expectations, and faces a tougher judge.. if she loses here children, the risk model can claim another successful prediction

171

the tool is meant to support human decision-making, not replace it.. nevertheless.. allegheny cty’s experiment is worth watching w a skeptical eye.. similar systems have been implemented recently in florida, la, nyc, oklahoma and oregon..

5 – the digital poorhouse

175

poverty in america is not invisible. we see it, and then we look away..t

mark rank: 51% of americans will spend at least a year below the poverty line between the ages of 20 and 65.. and yet we pretend that poverty is a puzzling aberration that happens only to a tiny minority of pathological people..t

cultural denial .. is a social process organized and supported by schooling govt, religion, media and other institutions..

when we pass (the man on the corner)…. it s because we have collectively convinced ourselves that there is nothing we can do for him.. when we failed to meet each other’s eyes as we passed, we signaled that , deep down, we know better. we could not make eye contact because we were enacting a cultural ritual of not-seeing,..t.. a semiconscious renunciation of our responsibility to each other.. our guilt, kindled because we perceived suffering and yet did nothing about it, made us look away.. we avoid not only the man on the corner.. but each other..

176

denial is exhausting and expensive. .. it contorts our physical geography as we build infra – suburbs, highways, private schools, and prisons .. that allow the professional middle class to actively avoid sharing the lives of poor and working  class..

poverty in america is actively denied by the way we define it: as falling below an *arbitrary income line at a single moment in time..

*an arbitrary man made ill

our public policy fixates on attributing blame for poverty rather than remedying its effects of abolishing its causes..

177

when poor and working people in the us become a politically viable force, relief institution and their techs of control shift to better facilitate cultural denial and to rationalize a brutal return to subserviency.. relief institutions are machines for undermining the collective power of poor and working class people, and for producing indifference in everyone else

178

this myopic focus on what’s new leads us to miss the important ways that digital tools are embedded in old systems of power and privilege..

in earlier chapter.. i provided an on the ground view of how new high tech tools are operating in social service programs across the country. it’s crucial to listen to those who are their primary targets; the stories they tell are diff than those told from the perspective of admin and analysts. now, i will zoom out to give a bird’s eye view of how these tools work together to create a shadow institution for regulating the poor.. t

179

1\ divert poor from public resources.. (indiana’s *eligibility automation).. when direct diversion fails.. moral narrative that criminalized most of the poor while providing life saving resources to a lucky handful

*too much ness

180

2\ classify and criminalize the poor (la) .. the  data of unhoused angelenos who receive no resources at all.. 21 500 people as of this writing – stay in the homeless management info system for 7 yrs.. lapd can access w/o warrant

181

when digital poorhouse simply bars access to public benefits.. as in indiana.. it is fairly easy to confront. but classification and criminalization work by including poor and working class in systems that limit their rights and deny their basic human needs.. doesn’t just exclude,.. ti sweeps millions of people into a system of control

3\ predict the future behavior of the poor (allegheny cty) .. prediction promises to produce hierarchies of worth and deservingness using stats and existing data instead of engaging human beings.. prediction unlike classification is intergenerational.. the impacts are thus exponential..

183

yesterday we experimented on the corpses of the poor; today we tinker w their futures

188

justice requires the possibility of redemption and the ability to start over..t

everyday.. equity

we measure human worth based only on the ability to earn a wage, and suffer in a world that undervalues care and community..

earn a living ness..

190

when automated decision making tools are not built to explicitly dismantle structural inequities, their speed and scale intensify them..

194

*liberty, **equity, ***inclusion

*krishna free will, **equity: everyone getting a go everyday, ***none of us if one of us

197

digital poorhouse was created in 70s.. the nation was asking difficult question: what is our obligation to each other in condition of ineq? how do we reward caregiving? the digital poorhouse reframed these big political dilemmas as mundane issues of efficiency and system engineering: how best match need to resource? how eliminate fraud and diver the ineligible?  how do we do most w the least money? allowed us to drop the bigger, more crucial convo

198

in 2012 economic ineq in us reached its highest level since 1928

200

when a very efficient tech is deployed against a despised outgroup in the absence of *strong human rights protections there is enormous potential for atrocity.. t

*gershenfeld sel

if there is to be an alternative, we must build it on purpose, brick by brick and byte by byte..

and everyday..  by each one of us.. eudaimoniative surplus

conclusion – dismantling the digital poorhouse

202

(1968 – on king calling to rid our nation and the world of poverty’.. and assassinated four day s later.. fbi, surveil, cointelpro, et al

204

the poor people’s campaign is one of our nation’s great unfinished journeys..

despite our unparalleled communications capabilities, we are in the midst of a violent retrenchment on equity and pluralism.. .. it will take more than high tech tweaks to bring down the institutions we have built to profile, police and punish the poor.. the most important step in dismantling the digital poorhouse is changing how we think, talk and feel about poverty..

not part\ial.. for (blank)’s sake… – as it could be

207

honkala: listening to the voices of those who have been told to be quiet and to disappear is incredibly important, strategic and vital..

begs a mech to listen to all the voices..not voiceless..  as it could be

208

(on orgs – cross gen et al.. led by poor facing difficulties attracting resources et al)..  the orgs led by the professional middle class on behalf of the poor, on the other hand, are more successful in attracting funding, progressive allies and public attention. but they are often disconnected form the radical analysis and boundless energy of poor and working class communities..

begs.. our data is idio-jargon/self-talk as data.. via 2 convos .. listening to and facilitating 7 bn curiosities .. everyday

on king’s bill of rights.. 1\ job  2\ min income  3\ house  4\ ed  5\ vote  6\ health

let’s simplify that.. so that 7 bn would resonate with it.. crave it.. today.. ie: authenticity and attachment.. maté basic needs

210

on ubi

let’s try ubi as temp placebo.. because if we don’t make money irrelevant (measuring transactions and validating people).. we’ll keep circling/cycling back around to poverty

212

gut check: 1\ does the tool increase the self determination and agency of the poor  2\ would the tool be tolerated if it was targeted at non poor people

yes – authenticity.. and yes – not tolerated.. craved.. ie: interconnectedness ..

we must begin dismantling the digital poorhouse.. it will require flexing our imagination and asking entirely diff kinds of questions..

*how would a data based system work if it was meant to encourage poor and working class people to use resources to meet their need in their own ways..t

*hlb via 2 convos that io dance.. as the day..

**what would decision making system that see poor people, families and neighborhoods as infinitely valuable and innovative look like?

**redefine decision making

***oath of non harm

***gershenfeld sel

214

the state doesn’t require a cop to kill a person

begs we go deeper (than we ever have).. to the roots of healing.. beyond blm..  cure violence.. poor people movement.. strike debt.. arab spring.. occupy.. whatever.. et al

215

addressing the digital poorhouse can help progressive social movements shift attention from ‘the policed’ to the processes of policing

how about we shift attention to cure ios city

in my most pessimistic moments, i fear that we are winning the fight against mass incarceration at just the historical moment when the digital poorhouse makes the physical institution of the prison less necessary

_________

from twitter convo after i’d read/tweeted..

@david82wilkins @EmilyK100 @PopTechWorks @monk51295 @zephoria @Livingstone_S @nathanjurgenson Have a look at this https://t.co/vk1nBfo9JV #theriseofthedigitalpoorhouse The threat AI poses, when employed without thorough thought, is frightening. Have you seen #IDanielBlake ? If so this depiction of reality doesn’t even touch the sides of whats to come #SocialWork

Original Tweet: https://twitter.com/AMLTaylor66/status/957680945934585857

I’ve blogged about my experience of it google.co.uk/amp/s/amltaylo…

i daniel blake (trailer in link to her blog)

__________

Virginia Eubanks (@PopTechWorks) tweeted at 3:35 PM on Wed, Feb 21, 2018:
This thread, y’all. This thread… https://t.co/AcU5NVOenG
(https://twitter.com/PopTechWorks/status/966441250646872065?ref_src=twcamp%5Eshare%7Ctwsrc%5Eandroid%7Ctwgr%5Egmail%7Ctwcon%5E7090%7Ctwterm%5E0)

by @MandyHenk

I love the potential of data and AI and algorithms. I am excited about what they can do. But I am also wondering about today’s 15 year old runaway. She deserves systems based on empathic humans making empathetic human decisions.

or one that lets her make decisions..

I got to control my story without having to engage with a computer system.

But there are folks who do really amazing social science that rounds out stories. Read Automating Inequality. Read Algorithms of Oppression by .

The tech world needs the voices of women like these, it needs to listen to social scientists asking us to put some breaks on and talk through this stuff.

Nothing about the world is inevitable..t

let’s infra for that..

complexity ness

_________

de-ice

Janus Cassandra  (@zenalbatross) tweeted at 6:24 AM – 26 Jun 2018 :
This is wild. All of us tech researchers are sitting here warning about the small, subtle ways algorithmic bias can support fascism — and meanwhile ICE has essentially created a button that prints “detain immigrant” when pressed https://t.co/kfbXJrOq6L (http://twitter.com/zenalbatross/status/1011586002257891328?s=17)

_________

algorithm

algos of oppression

weapons of math

poverty

equity – everyone getting a go everyday

as it could be

ie: hlb via 2 convos that io dance.. as the day..[aka: not part\ial.. for (blank)’s sake…]..  a nother way

__________

Advertisements