Jekyll2023-10-25T20:13:45+00:00https://tiwong.github.io/feed.xmlWONG Tian An 黄天安Hello worldSome things Southeast Asia2022-09-27T11:50:45+00:002022-09-27T11:50:45+00:00https://tiwong.github.io/theory/2022/09/27/some-things-sea<p>Besides working on moving <a href="https://tiwong.github.io/theory/2021/10/15/insurgent-sea-call.html">this volume</a> along, here’s some updates about what (else) I’ve been doing:</p>
<ul>
<li>A <a href="https://get-sea.org/past-courses/">GETSEA minicourse</a> on Anarchism and Southeast Asia! <a href="/files/Anarchism_Southeast_Asia_Syllabus.pdf">See the syllabus</a>.</li>
<li><a href="https://criticalasianstudies.org/commentary/2022/4/17/notes-from-the-field-wong-tian-an-should-southeast-asian-studies-exist-field-notes-from-an-interloper">This post</a> in Critical Asian Studies on whether Southeast Asian Studies (in the West) should exist.</li>
<li>I was also for some reason asked to serve on the executive committee for the University of Michigan Center of Southeast Asian Studies (<a href="https://ii.umich.edu/cseas/people/executive-committee.html">CSEAS</a>), so how about that?</li>
<li>And hopefully another minicourse next year on abolition!</li>
</ul>Besides working on moving this volume along, here’s some updates about what (else) I’ve been doing:Manuscript: An Asian American Theology of Liberation2022-01-29T11:50:45+00:002022-01-29T11:50:45+00:00https://tiwong.github.io/theory/2022/01/29/manuscript-liberation-theology-unbound<p>I’ve gotten tired of waiting to get the book published. And anyway, I had intended the text to be freely accessible from the beginning. So here is a full manuscript/draft of An Asian American Theology of Liberation. Please send me comments and feedback! UPDATE: <a href="https://www.fulcrum.org/concern/monographs/hd76s301r">Here is the book</a>, open access!</p>
<p>Also, for some background, see <a href="https://tiwong.github.io/theory/2019/03/21/to-the-archives.html">this post</a>.</p>I’ve gotten tired of waiting to get the book published. And anyway, I had intended the text to be freely accessible from the beginning. So here is a full manuscript/draft of An Asian American Theology of Liberation. Please send me comments and feedback! UPDATE: Here is the book, open access!Call for Contributors: Insurgent Southeast Asia2021-10-15T11:50:45+00:002021-10-15T11:50:45+00:00https://tiwong.github.io/theory/2021/10/15/insurgent-sea-call<p><img src="/files/Call_ISEA.jpg" alt="Call" title="Call for Contributors: Insurgent Southeast Asia" /></p>
<p><em>Where will this be published?</em> At the moment, the Strategic Information and Research Development Centre (SIRD) at <a href="https://www.gerakbudaya.com/SIRD">Gerakbudaya</a> has expressed interest in publishing the volume. There is also a possibility of co-publishing with another academic publisher, but all this is contingent upon having a strong manuscript. This is where <em>you</em> come in!</p>
<p><em>What are you looking for?</em> Anything that broadly fits the bill in the call above. If you’re unsure or want to workshop an idea, write to me! At a basic level, I am hoping for:</p>
<p>Projected sections to include at least one contributor for each of the following countries:</p>
<ol>
<li>Indonesia (e.g., protests against the omnibus law)</li>
<li>Malaysia (e.g., protests against the Pakatan Nasional backdoor government; cleaning worker and contract doctors strikes; student protests against sexual harassment)</li>
<li>Myanmar (e.g., the ongoing National Unity Government led revolution)</li>
<li>Philippines (e.g., protests against the Duterte government; Chinese encroachment in the South China Sea; land struggles of the indigenous Lumad)</li>
<li>Singapore (e.g., climate activists and independent journalists fighting against the continued suppression of dissent; vaccine nationalism; the Raffles Must Fall movement)</li>
<li>Thailand (e.g., protests against the monarchy in Thailand)</li>
<li>Vietnam</li>
</ol>
<p>Also, from a comparative aspect, potential transnational topics include:</p>
<ol start="8">
<li> Critical analysis of the Milk Tea Alliance from an anti-statist perspective</li>
<li> The colonial logics and logistics of palm oil industry in West Papua embedded in global financial networks, including Malaysia, Singapore, and USA</li>
<li> Police and prison abolition as a means of transnational organising/abolition outside a Eurocentric frame </li></ol>
<p><br />
…plus an interview with James Scott, Yale University.</p>
<p><em>How strict are the deadlines?</em> They’re not, so if you need additional time but want to be kept in consideration, again just write to me! But most importantly, get in touch so that I am aware that you are interested.</p>Labour, etc.2021-09-17T11:50:45+00:002021-09-17T11:50:45+00:00https://tiwong.github.io/theory/2021/09/17/labour-etc<p>It’s been awhile! Here is a list of links to things I’ve written for the public in the meanwhile, mainly dealing with labour in the academy, but not entirely.</p>
<ul>
<li><a href="https://wolverzine.org/wp-content/uploads/2022/02/Transformations_Vol1.pdf">Working Together, Attending Class</a>, contribution to <em>Wolverzine</em>, a zine about students as partners</li>
<li><a href="https://www.timeshighereducation.com/depth/university-really-community">Capital Gains</a>, contribution to “Is the university really a community?” <em>Times Higher Education</em></li>
<li><a href="https://www.timeshighereducation.com/blog/weve-seen-chair-now-its-time-adjunct">We’ve seen ‘The Chair’. Now it’s time for ‘The Adjunct’</a>, <em>Times Higher Education</em></li>
<li><a href="https://dearbornhub.net/?p=1018">Reflections on the Inside-Out Program training</a>, <em>UM-Dearborn Hub for Teaching and Learning Resources</em></li>
<li><a href="https://blogs.ams.org/inclusionexclusion/2021/02/01/the-mathjob-market-is-bad-but-what-else-is-new-a-2020-retrospective/">The mathjob market is bad. But what else is new? A 2020 retrospective</a>, <em>Inclusion/Exclusion</em></li>
<li><a href="https://blogs.ams.org/inclusionexclusion/2020/06/30/math-in-pandemic-and-precarity/">Math, in pandemic and precarity</a>, <i>Inclusion/Exclusion</i></li>
<li><a href="https://www.timeshighereducation.com/opinion/us-hiring-freeze-leaving-contingent-faculty-out-cold">The US hiring freeze is leaving contingent faculty out in the cold</a>, <i>Times Higher Education</i></li>
<li><a href="https://blogs.ams.org/inclusionexclusion/2020/01/31/can-mathematics-be-antiracist/">Can mathematics be antiracist?</a>, <em>Inclusion/Exclusion</em></li>
</ul>
<p>Also, <a href="https://www.gazettenet.com/How-tenants-organized-around-a-rent-increase-35926429">in the news</a>.</p>It’s been awhile! Here is a list of links to things I’ve written for the public in the meanwhile, mainly dealing with labour in the academy, but not entirely. Working Together, Attending Class, contribution to Wolverzine, a zine about students as partners Capital Gains, contribution to “Is the university really a community?” Times Higher Education We’ve seen ‘The Chair’. Now it’s time for ‘The Adjunct’, Times Higher Education Reflections on the Inside-Out Program training, UM-Dearborn Hub for Teaching and Learning Resources The mathjob market is bad. But what else is new? A 2020 retrospective, Inclusion/Exclusion Math, in pandemic and precarity, Inclusion/Exclusion The US hiring freeze is leaving contingent faculty out in the cold, Times Higher Education Can mathematics be antiracist?, Inclusion/ExclusionTo the archives!2019-03-21T11:50:45+00:002019-03-21T11:50:45+00:00https://tiwong.github.io/theory/2019/03/21/to-the-archives<p>Did you know Asian American liberation theology was a thing in the past? Well, it was. Alongside Black liberation theology, Latin American liberation theology, and white feminist liberation theology. What’s left of it is hidden in dusty archives, most of it in the <a href="https://oac.cdlib.org/findaid/ark:/13030/kt75804087/entire_text/">Graduate Theological Union archives</a>. I’ve managed to get some of them digitised. The second reader is done by myself, with thanks to the Vancouver School of Theology, and the rest are thanks to D. Stiver at the GTU, posted with permission of Roy I. Sano.</p>
<ul>
<li><a href="/files/SanoReader1973.pdf">Amerasian Theology of Liberation: A Reader (1973)</a></li>
<li><a href="/files/SanoAutobiographical1973.pdf">Autobiographical Reflections for Amerasian Theology of Liberation (1973)</a></li>
<li><a href="/files/SanoProceedings1975.pdf">East Asian and Amerasian Liberation: Second Conference on East Asian and Amerasian Theology (1975)</a></li>
<li><a href="/files/SanoReader1976.pdf">The Theologies of Asian Americans and Pacific Peoples: A Reader (1976)</a> (warning: 62 MB!)</li>
</ul>
<p>For some context, check out:</p>
<ul>
<li><a href="http://camla.org/wp-content/uploads/2017/05/CAM-Roots-Zine-2017.pdf">Roots: Asian American Movements in Los Angeles</a>, the zine accompanying the “Roots: Asian American Movements 1968-80s” exhibit at the Chinese American Musuem, LA in 2017. The name harkens to the out of print 1971 <em>Roots: An Asian American Reader</em> edited by Amy Tachiki</li>
<li><a href="https://en.wikipedia.org/wiki/Gidra_(newspaper)">Gidra</a>, the revolutionary newspaper-magazine, from 1969 to 1974 is available at the <a href="http://ddr.densho.org/ddr/densho/297/">Densho Digital Repository</a>.</li>
<li><a href="https://www.cia.gov/library/readingroom/docs/CIA-RDP97R00694R000600050001-9.pdf">Liberation Theology: Religion, Revolution, Reform</a> April 1986 declassified CIA (Directorate of Intelligence) research paper</li>
</ul>
<p>Some books on the Asian American history of that era:</p>
<ul>
<li><em>Serve the People: Making Asian America in the Long Sixties,</em> Karen Ishizuka</li>
<li><em>Legacy to Liberation: Politics and Culture of Revolutionary Asian Pacific America,</em> Fred Ho (ed)</li>
</ul>
<p>and for Asian American histories at large:</p>
<ul>
<li><em>American History Unbound: Asians and Pacific Islanders,</em> Gary Y. Okihiro</li>
<li><em>Asian American Dreams: The Emergence of an American People</em>, Helen Zia</li>
<li><em>Strangers from a Different Shore: A History of Asian Americans,</em> Ronald Takaki</li>
<li><em>The Making of Asian America: A History</em>, Erika Lee</li>
</ul>
<p>Finally, some things I’ve written:</p>
<ul>
<li><a href="https://www.inheritancemag.com/article/asian-american-liberation-theology">Asian American Liberation Theology: A past we never knew was ours</a>, <em>Inheritance</em></li>
<li><a href="https://www.inheritancemag.com/article/asian-american-liberation-theology-now">Asian American Liberation Theology Now!</a>, <em>Inheritance</em></li>
<li><a href="https://sojo.net/articles/why-we-need-theology-protest">Why We Need A Theology of Protest</a>, <em>Sojourners</em></li>
<li><a href="https://sojo.net/articles/poem-love-letter-hawaii-puerto-rico-and-hong-kong">Love Letter to Hawai’i, Puerto Rico, and Hong Kong</a>, <em>Sojourners</em></li>
<li><a href="https://www.divergingmag.com/unsettling-asian-american-theology/">Unsettling Asian American Theology</a>, <em>Diverging</em></li>
<li><a href="http://reformedmargins.com/whos-afraid-of-liberation-theology-a-call-to-liberation/">Who’s Afraid of Liberation Theology? A Call to Liberation</a>, <em>Reformed Margins</em></li>
<li><a href="/files/WongAAR_final.pdf">Rearticulating an Asian American Theology of Liberation</a>, <em>American Academy of Religion</em></li>
<li><a href="/files/GidraWong.pdf">Asian American liberation theology: Searching for an Asian radical tradition</a>, <em>Gidra</em> March 2020 Issue on Asian Americans and Blackness in America</li>
<li><a href="https://www.divergingmag.com/lent-week-6/">A Pandemic Passover: An invitation to Asian American Liberative Praxis</a>, Diverging</li>
<li><a href="https://www.inheritancemag.com/stories/asians-in-black-riots">Asians in Black Riots</a>, <em>Inheritance</em></li>
<li><a href="https://lausan.hk/2020/black-liberation-is-asian-liberation/">Black liberation is Asian Liberation</a> <em>Lausan</em></li>
</ul>
<p>Now here’s where you come in. I’ve put together a User’s Guide below as a template for you to kick off discussion and build an Asian American theology of liberation for the next generation. It is as much yours as it is mine. Have at it!</p>
<ul>
<li><em>A User’s Guide to Asian American Theologies of Liberation</em> (alpha version, 2019) [<a href="/files/UsersGuide.docx">docx</a>] [<a href="/files/UsersGuide.pdf">pdf</a>]</li>
</ul>
<p>UPDATE: <a href="https://tiwong.github.io/theory/2022/01/29/manuscript-liberation-theology-unbound.html">See this post</a> for my manuscript of Liberation Theology Unbound: An Asian American theology of liberation.</p>Did you know Asian American liberation theology was a thing in the past? Well, it was. Alongside Black liberation theology, Latin American liberation theology, and white feminist liberation theology. What’s left of it is hidden in dusty archives, most of it in the Graduate Theological Union archives. I’ve managed to get some of them digitised. The second reader is done by myself, with thanks to the Vancouver School of Theology, and the rest are thanks to D. Stiver at the GTU, posted with permission of Roy I. Sano.To Talk Like Them2018-06-25T11:50:45+00:002018-06-25T11:50:45+00:00https://tiwong.github.io/theory/2018/06/25/to-talk-like-them<p>I wrote a thing! Read it <a href="https://fieldworking.net/2018/06/21/to-talk-like-them/">here</a> or <a href="https://footnotesblog.com/2018/06/25/to-talk-like-them/">here</a>. Many thanks to Jessica for holding space!</p>I wrote a thing! Read it here or here. Many thanks to Jessica for holding space!Warm regards2017-10-16T11:50:45+00:002017-10-16T11:50:45+00:00https://tiwong.github.io/climate/2017/10/16/warm-regards<p>I read somewhere, with regards to the Trump administration, that experts on authoritarianism warn that you should keep a list of the changes that are happening so as not to forget the way things were, that the new normal is <em>not</em> normal. The erosion of civil liberties is exactly that: a slow chipping away at your fundamental rights, so that by the time you realise what is happening, it is already too late.</p>
<p>There’s a chance what’s going on with climate change is already too late, but only time will tell. This page maintains a list of phenomena around the world, related to global warming or climate change, and will be continually updated.</p>
<p>As a scientist, I feel it’s my duty encourage you to dig through to the original source when possible. Almost always, reporting about scientific progress is exaggerated and distorted. To start, here’s a <a href="https://www.nytimes.com/interactive/2017/climate/what-is-climate-change.html">primer</a> on climate change by the NYT.</p>
<ul>
<li>
<p>7/2018 <a href="https://teachclimatescience.wordpress.com/">The Teacher-Friendly Guide™ to Climate Change</a> looks like a great resource for teaching about climate change at the high school level. The e-book is free to download!</p>
</li>
<li>
<p>6/2018 <a href="https://www.researchgate.net/publication/311844520_Carbon_dioxide_toxicity_and_climate_change_a_serious_unapprehended_risk_for_human_health">Carbon dioxide toxicity and climate change.</a> “The toxicity of CO2 for breathing has been well defined for high concentrations but it remains effectively unknown what level will compromise human health when individuals are perpetually exposed for their lifetime. There is evidence from the few studies of long-term low-level exposure that permanent exposure, to CO2 levels predicted by the end of the century, will have significant effects on humans.”</p>
</li>
<li>
<p>5/2018 <a href="https://www.theguardian.com/cities/ng-interactive/2017/nov/03/three-degree-world-cities-drowned-global-warming">The third-degree world: cities that will be drowned by global warming</a></p>
</li>
<li>
<p>5/2018 <a href="http://www.sciencemag.org/news/2016/11/watch-air-pollution-flow-across-planet-real-time">AirVisual Earth.</a> Watch air pollution flow across the planet in real time. Also, a Nature article on <a href="https://www.nature.com/articles/s41558-018-0141-x">the carbon footprint of global tourism.</a> and a visualisation of the fact that <a href="https://theatln.tc/2G7BE0w">every Google search results in CO2 emissions.</a></p>
</li>
<li>
<p>5/2018 <a href="https://www.scientificamerican.com/article/shock-and-thaw-alaskan-sea-ice-just-took-a-steep-unprecedented-dive/">Alaskan Sea Ice</a></p>
</li>
<li>
<p>4/2018 <a href="https://qz.com/1249126/a-new-study-on-increased-snowfall-in-antarctica-shows-the-dramatic-pace-of-climate-change/">Antartica is turning into a snow globe.</a> “Warmer temperatures mean more moisture in the air, which creates better conditions for snow over Antartica. So really, this is a sign of the same climate problems causing droughts, storms, and floods.” Further north, the <a href="https://www.scientificamerican.com/article/slow-motion-ocean-atlantics-circulation-is-weakest-in-1-600-years/">Atlantic’s circulation is weakest in 1600 years.</a> “If hemisphere-spanning currents are slowing, greater flooding and extreme weather could be at hand.” Also, <a href="https://www.scientificamerican.com/article/ocean-heat-waves-are-getting-worse/">ocean heat waves are getting worse</a>.</p>
</li>
<li>
<p>3/2018 More on Greenland, <a href="https://www.scientificamerican.com/article/greenland-is-melting-faster-than-any-time-in-the-last-400-years/">melting faster than any time in the last 400 years.</a> “Melting in western Greenland is now nearly double what it was at the end of the 19th century, research suggests.)</p>
</li>
<li>
<p>1/2018 <a href="https://www.nytimes.com/interactive/2018/01/18/climate/hottest-year-2017.html">2017 was one of the hottest years on record. And that was without El Niño.
</a> The title about says it all. Without El Niño, meaning that the world is now experiencing a weak La Niña, so average mean temperatures aren’t expected to break records again until the next El Niño. Which is soon enough.</p>
</li>
<li>
<p>11/2017 <a href="https://www.vox.com/energy-and-environment/2017/11/21/16685876/climate-change-clock-ticker-global-warming-gif">Watch climate change in real time</a> with these trackers. These tickers show the unrelenting rise in global temperatures and carbon dioxide.</p>
</li>
<li>
<p>10/2017 <a href="https://digiconomist.net/bitcoin-energy-consumption">Bitcoin energy consumption.</a> Bitcoin’s current estimated annual electricity consumption is around 33TWh, just a little more than Denmark, and a little less than Belarus. “Bitcoin’s biggest problem is not even its massive energy consumption, but that the network is mostly fueled by coal-fired power plants in China. Coal-based electricity is available at very low rates in this country. Even with a conservative emission factor, this results in an extreme carbon footprint for each unique Bitcoin transaction.”</p>
</li>
<li>
<p>10/2017 <a href="https://www.bloomberg.com/news/articles/2017-10-06/there-s-a-climate-change-bomb-under-your-feet">There’s a climate bomb under your feet</a>. “What they found, published yesterday in the journal Science, may mean the accelerating catastrophe of global warming has been fueled in part by warm dirt. As the Earth heats up, microbes in the soil accelerate the breakdown of organic materials and move on to others that may have once been ignored, each time releasing carbon dioxide into the atmosphere.”</p>
</li>
<li>
<p>9/2017 <a href="http://www.politico.com/agenda/story/2017/09/13/food-nutrients-carbon-dioxide-000511">The great nutrient collapse</a>. “Every leaf and every grass blade on earth makes more and more sugars as CO2 levels keep rising. We are witnessing the greatest injection of carbohydrates into the biosphere in human history―[an] injection that dilutes other nutrients in our food supply.”</p>
</li>
<li>
<p>8/2017 <a href="https://earthobservatory.nasa.gov/blogs/earthmatters/2017/08/10/roundtable-the-greenland-wildfire/">Wildfire in Greenland</a>. For over two weeks, 40 miles from the ice sheet. UPDATE: <a href="https://www.theatlantic.com/science/archive/2017/11/the-zombie-diseases-of-climate-change/544274/">Reporting</a> on the affects of melting permafrost—including in Greenland—re-animating ancient viruses and bacteria.</p>
</li>
<li>
<p>7/2017 <a href="https://www.nasa.gov/feature/goddard/2017/massive-iceberg-breaks-off-from-antarctica">The Larsen C shelf</a>. A floating platform of glacial ice on the east side of the Antarctic Peninsula, is the fourth largest ice shelf ringing Earth’s southernmost continent. In 2014, a crack that had been slowly growing into the ice shelf for decades suddenly started to spread northwards, creating the nascent iceberg. Now that the close to 2,240 square-mile (5,800 square kilometers) chunk of ice has broken away, the Larsen C shelf area has shrunk by approximately 10 percent. [14.11.2017 UPDATE: See ongoing reports from NASA’s <a href="https://earthobservatory.nasa.gov/blogs/fromthefield/category/operation-icebridge-2017/">Operation Icebridge</a>]</p>
</li>
<li>
<p>10/2016 <a href="https://www.outsideonline.com/2112086/obituary-great-barrier-reef-25-million-bc-2016">Obituary: Great Barrier Reef</a>. “In 1981, the same year that UNESCO designated the reef a World Heritage Site and called it “the most impressive marine area in the world,” it experienced its first mass-bleaching incident. Corals derive their astonishing colors, and much of their nourishment, from symbiotic algae that live on their surfaces. The algae photosynthesize and make sugars, which the corals feed on. But when temperatures rise too high, the algae produce too much oxygen, which is toxic in high concentrations, and the corals must eject their algae to survive. Without the algae, the corals turn bone white and begin to starve. If water temperatures soon return to normal, the corals can recruit new algae and recover, but if not, they will die in months. In 1981, water temperatures soared, two-thirds of the coral in the inner portions of the reef bleached, and scientists began to suspect that climate change threatened coral reefs in ways that no marine park could prevent.”</p>
</li>
<li>
<p>3/2010 <a href="http://www.pnas.org/content/107/21/9552.abstract">An adaptability limit to climate change due to heat stress</a>. “One implication is that recent estimates of the costs of unmitigated climate change are too low unless the range of possible warming can somehow be narrowed. Heat stress also may help explain trends in the mammalian fossil record.”</p>
</li>
</ul>I read somewhere, with regards to the Trump administration, that experts on authoritarianism warn that you should keep a list of the changes that are happening so as not to forget the way things were, that the new normal is not normal. The erosion of civil liberties is exactly that: a slow chipping away at your fundamental rights, so that by the time you realise what is happening, it is already too late.How to use math for evil IV: Academia2017-10-06T11:50:45+00:002017-10-06T11:50:45+00:00https://tiwong.github.io/data/2017/10/06/how-to-use-math-for-evil-4<p><em>This is an expanded version of the general audience talk I’m giving at the Science on Tap–formerly known as Drunk on Science, which is much cooler but probably less professional–in Pune, the brainchild of the illustrious Anoop Mahajan, with craft beer sponsored by Great State Ale Works! See <a href="https://tiwong.github.io/data/2017/09/28/how-to-use-math-for-evil-1.html">Part 1</a>, <a href="https://tiwong.github.io/data/2017/09/30/how-to-use-math-for-evil-2.html">Part 2</a>, <a href="https://tiwong.github.io/data/2017/10/03/how-to-use-math-for-evil-3.html">Part 3</a>, <a href="https://tiwong.github.io/data/2017/10/06/how-to-use-math-for-evil-4.html">Part 4</a></em></p>
<p>In the final part of this series, we’ll look at how mathematics operates within academia. I won’t be touching on this in the talk as it’s more of a special interest topic; also this is possibly professionally risky as someone far away from being a tenured professor. But the issues raised below are highly pertinent today, and it’s important to be aware of the waters we swim in.</p>
<p>What we’ll look at in this post does not directly involve the content of pure, or even applied mathematics, but rather the culture of modern mathematics, the power structures that support it and vice versa.</p>
<h4 id="an-ethnography">An ethnography</h4>
<p>A well-known ethnography of a biological institute was conducted in the late 70s by sociologists Bruno Latour and Steve Woolgar. <a href="https://en.wikipedia.org/wiki/Laboratory_Life"><em>Laboratory Life: The Construction of Scientific Facts</em></a> studied the social dimensions inherent in the production of scientific knowledge.</p>
<blockquote>
<p>“Whereas we now have fairly detailed knowledge of the myths and circumcision rituals of exotic tribes, we remain relatively ignorant of the details of equivalent activity among tribes of scientists, whose work is commonly heralded as having startling or, at least, extremely significant effects on our civilisation.” – Bruno Latour (p.17)</p>
</blockquote>
<p>By and large, the tools of anthropological study are often applied in the direction that power is distributed. Of course, it’s quite well known that it is the coloniser that gets to study the colonised, and inverting the lens of study often dredges up complaints like ‘reverse racism.’ This, like notions of white fragility when one’s position of privilege is exposed, reflects upon the obvious lack of empathy inherent in hegemonic structures.</p>
<p>The act of criticising the academy’s complicity with oppressive systems is not new, but its persistence makes it imperative for us to continually consider its failings, to speak truth to power, and of course, acknowledge our own complicity wherever we witness it at work. Intersectional theories—of the non-mathematical kind!—tell us that working in the ivory tower, we necessarily operate within a multiplicity of oppressive systems at any given point in time. The fact that these often resonate with contemporary struggles locates the academy as another site of struggle and potential solidarity. Many of the issues discussed will not be particular to mathematics, but I will discuss them in the context of mathematics.</p>
<p>The ethnography, that I mean to describe briefly, is my own witness of mathematics culture as observed in various institutions primarily in the US, Germany, and India. Most of all my two-month stay in the mathematics institute at Oberwolfach, where every week saw a different group of forty-odd mathematicians convene to discuss the most recent advances and problems in their field of specialisation.</p>
<h4 id="math-is-awesome-and-math-culture-is-terrible">“Math is awesome and math culture is terrible”</h4>
<p>As such, the first point here is the <em>jet-setting academic</em>. Each week a different group of experts is summoned from across the world, which certainly requires a good deal of air travel. These are almost always sponsored by faculty grants or by the institution itself, which receives a fair amount of government funding. At most, this large, sustained carbon footprint is acknowledged—and I implicate myself in this—with a somewhat apologetic tone, but also with a disavowal of responsibility by invoking a perceived inevitability. The reality of one’s socio-ecological responsibility is essentially waived by a fantasy that there really isn’t anything that can be done about it.</p>
<p>As we shall see, the most pernicious stain on pure mathematics research is not that it could be actively used ‘for evil,’ but rather the widespread and deeply entrenched <em>apathy</em>. As the saying goes, “All that is necessary for evil to triumph is for good men to do nothing.”</p>
<p>The second point, is what is derisively called <em>the cult of genius</em>. While certainly there are those who are more naturally gifted in mathematics, the cult of genius that is pervasive in mathematics—not to mention the ageism inherent in the Fields Medal which only serves to exacerbate this notion—has a damaging effect on ordinary students of mathematics. This stereotype goes against the growth mindset and resilience that ought instead to be nurtured in students. <a href="https://www.inc.com/kimberly-weisul/how-harvey-mudd-college-achieved-gender-parity-computer-science-engineering-physics.html">According</a> to Maria Klawe, President of Harvey Mudd College,</p>
<blockquote>
<p>“In disciplines that say it’s about innate ability, there are fewer women.”</p>
</blockquote>
<p>But not only are there fewer women, but it stand to reason from the more general phenomenon that success in college is <a href="https://www.vox.com/2017/9/11/16270316/college-mobility-culture">biased against minority groups</a>, that those who largely succeed are white males. And this is provably true.</p>
<h4 id="mathematics-from-the-margins">Mathematics from the Margins</h4>
<p>So let’s talk about the <em>cis-het white male</em>. Though mathematicians as a whole like to view themselves as a left-leaning community, it remains that the majority of mathematicians, alive or dead, are white male and cisgender heterosexual. This is particularly important, because the old white boys club, left to its own devices, will perpetuate itself, as demonstrated in <a href="https://www.scientificamerican.com/article/male-scientists-share-more-but-only-with-other-men/">this informal experiment</a>. And it is only a small saving grace that mathematics does not require field trips, which is a hotbed of sexual harassment by senior male professors, as it is slowly being <a href="http://www.sciencemag.org/news/2017/10/disturbing-allegations-sexual-harassment-antarctica-leveled-noted-scientist">made known</a> and <a href="https://www.wired.com/2016/12/can-build-calendar-sexual-harassment-stories-science/">documented</a>.</p>
<p>Many inroads have been made by groups such as the Association for Women in Mathematics (AWM), but much more work remains to be done for those in ethnic and sexual minorities. Piper Harron, who has been voicing out against the issues she has faced as a black woman in mathematics has been met with much opposition, but has been supported on the new American Mathematical Society (AMS) blog <a href="https://blogs.ams.org/inclusionexclusion">Inclusion/Exclusion</a>. I encourage you to read everything she has written on the topic on <a href="http://www.theliberatedmathematician.com/blog/">her blog</a>. From the introduction to her doctoral dissertation:</p>
<blockquote>
<p>“Respected research math is dominated by men of a certain attitude. Even allowing for individual variation, there is still a tendency towards an oppressive atmosphere, which is carefully maintained and even championed by those who find it conducive to success. […] The problem was not individuals, but a system of self-preservation that, from the outside, feels like a long string of betrayals, some big, some small, perpetrated by your only support system.” —Piper Harron</p>
</blockquote>
<p>On the other hand <a href="https://lgbtmath.org">Spectra</a>, an association for LGBT+ mathematicians, was recently founded to provide recognition and community for gender and sexual minority mathematicians. Though, as noted by its co-founder Mike Hill, gender is performative, and in mathematical spaces it is rarely discussed. This comes back to the point of apathy: mathematics pretends to be color-blind and gender-blind, but the position of not taking a position is never neutral. Indeed, the pretension of neutrality simply serves to reproduce what Danny Martin calls the ‘racial hierarchy of mathematics education.’</p>
<p>The stereotype of asians being predominantly good at mathematics is part of a larger problem that is <em>the model minority myth</em>. While some might argue it is a positive stereotype, it is in fact damaging to asian americans as a community, being held to a higher standard when their innate ability to succeed in mathematics is no better or worse than anyone else’s. Rather, it is important to recognise that the model minority myth is a pillar supporting anti-black racism, and is historically facilitated by US immigration policies favouring wealthy, highly educated elites from asia to immigrate. Indeed, much of the racial representation that we see in the US mathematical community is in fact internationally ‘sourced’, and as such, professional mathematicians home-grown in the US remain sorely lacking in diversity. Through the model minority myth we can locate whiteness: non-asian students of colour are judged as inferior to their white classmates, whereas asian students who do even better than white students are explained away by the stereotype. So whiteness is the norm in mathematics.</p>
<p>Piper sums it up thus: <a href="http://www.thehindu.com/thread/arts-culture-society/article9022211.ece">Math is awesome and math culture is terrible</a>. We can also take a hint from Linda Nochlin, whose 1971 essay on <a href="http://www.artnews.com/2015/05/30/why-have-there-been-no-great-women-artists/"><em>Why Have There Been No Great Women Artists?</em></a> is considered the founding of feminist art history. It is worth quoting her at length, thinking about women mathematicians in place of artists:</p>
<blockquote>
<p>“But in actuality, as we all know, things as they are and as they have been, in the arts as in a hundred other areas, are stultifying, oppressive and discouraging to all those, women among them, who did not have the good fortune to be born white, preferably middle class and, above all, male. The fault, dear brothers, lies not in our stars, our hormones, our menstrual cycles or our empty internal spaces, but in our institutions and our education—education understood to include everything that happens to us from the moment we enter this world of meaningful symbols, signs and signals. The miracle is, in fact, that given the overwhelming odds against women, or blacks, that so many of both have managed to achieve so much sheer excellence, in those bailiwicks of white masculine prerogative like science, politics or the arts.” — Linda Nochlin</p>
</blockquote>
<p>Nochlin goes on to take aim at the notion of Genius, or the ‘myth of the Great Artist,’ and examines the social conditions for producing art (or for us, mathematics), showing that the development of artists and their art is ‘mediated and determined by specific and definable social institutions’.</p>
<h4 id="the-adjunctification-of-the-academy">The Adjunctification of the Academy</h4>
<p>Implicitly, what I’ve described above pertains mainly to research mathematics, or mathematics performed by tenured or tenure-track faculty. As a proportion of those who teach mathematics in the college classroom, this is steadily decreasing, giving way to a growing army of contingent labor in the US academic system. Adjunct instructors are contract workers with little access to labour unions, health insurance, and proper workspaces. They are paid at a far lower rate than their faculty counterparts, and so cost universities much less to hire. For example, my own PhD advisor was convinced that since his retirement, his tenured position would be divided into several adjunct positions instead.</p>
<p>The adjunctification of academia exacerbates the labor struggle even of knowledge workers, creating a class difference within the faculty lounge, not just the classroom. Indeed, while adjuncting as a graduate student, at the end of every semester I received a letter saying that there was no guarantee that my position would be renewed, though I was assured by the chair that this was only a formality. For this reason, my time on the adjunct track was relatively protected, and as a graduate student it was understood that this position was only temporary. But for many others who have already earned doctorates but unable to land a tenure track get relegated to the ‘adjunct track’, a sort of eternal limbo. More extreme cases have grabbed the headlines such as <a href="https://www.theguardian.com/us-news/2017/sep/28/adjunct-professors-homeless-sex-work-academia-poverty">Facing poverty, academics turn to sex work and sleeping in cars</a>.</p>
<p>Adjunct labor is itself a product of the university as a capitalist institution producing far more aspiring academics than the academy itself can sustain. In turn the increasing number of trained intellectuals being subjected to oppressive labor conditions simply increase the likelihood that these intellectuals will become <a href="https://www.nytimes.com/2017/09/30/opinion/sunday/adjunct-professors-politics.html">radicalised against the academy</a>, which of course will find resistance by the host institutions themselves, which are inherently conservative. And all this not to mention the effect this has on <a href="https://www.theatlantic.com/education/archive/2015/05/the-cost-of-an-adjunct/394091/">the students that the adjuncts teach</a>. The conditions under which adjuncts labour under necessarily reduces the quality of their teaching, though this weakness must be hidden away, say, in the form of grade inflation, in order to preserve the adjunct’s job.</p>
<h4 id="the-psychological-brutality-of-the-post-doctoral-system">“The psychological brutality of the post-doctoral system”</h4>
<p>This precarity also extends to other forms of temporary labor in the academic workforce, even if slightly more respectable. The plight of postdocs in a publish-or-perish environment has brought to light issues of mental health in academia, for example a <a href="https://www.theguardian.com/science/head-quarters/2017/aug/10/the-human-cost-of-the-pressures-of-postdoctoral-research">particular case of suicide</a> which led to controversy surrounding a subsequent physics publication. An excerpt of the acknowledgements of Oliver Roston, in memory of his former colleague Francis Dolan:</p>
<blockquote>
<p>“I am firmly of the conviction that the psychological brutality of the post-doctoral system played a strong underlying role in Francis’ death. I would like to take this opportunity, should anyone be listening, to urge those within academia in roles of leadership to do far more to protect members of the community suffering from mental health problems, particularly during the most vulnerable stages of their careers.”</p>
</blockquote>
<p>The current trial-by-fire of the itinerant postdoc before going to bat for a tenure-track faculty position can be costly, in terms of mental health and emotional stability. Because of the nature of the acknowledgment, the paper was declined to be published by two journals, though the physics contained in the paper was up to standard, and finally accepted by a third journal. Moreover, a postdoc is no guarantee for a tenure job, and studies show it is worthless outside of academia](http://www.sciencemag.org/careers/2017/01/price-doing-postdoc). Quoting Julia Lane, an economist at New York University, “a postdoc is essentially high-quality cheap labor for the machine that is modern-day science.”</p>
<p>A side note on the publishing industry, which profits off of the backs of scientists, as it were: “ Scientists create work under their own direction – funded largely by governments – and give it to publishers for free; the publisher pays scientific editors who judge whether the work is worth publishing and check its grammar, but the bulk of the editorial burden – checking the scientific validity and evaluating the experiments, a process known as peer review – is done by working scientists on a volunteer basis. The publishers then sell the product back to government-funded institutional and university libraries, to be read by scientists – who, in a collective sense, created the product in the first place.” <a href="https://www.theguardian.com/science/2017/jun/27/profitable-business-scientific-publishing-bad-for-science">link</a> In the words of Dr Neal Young of the National Institutes of Health:</p>
<blockquote>
<p>“We scientists have not given a lot of thought to the water we’re swimming in.” —Neal Young</p>
</blockquote>
<h4 id="the-secret-weapon-of-cultural-imperialism">The secret weapon of cultural imperialism</h4>
<p>So far we’ve mainly looked at the structures surrounding the production of mathematics. Let’s circle back to think a little more about the reproduction of mathematics, namely, mathematics education. Now, there is a certain divide between mathematics education and mathematics research. For example, studies of mathematics education is largely focused on the grade school level, and scrutiny fades as we come closer to graduate-level mathematics. The result is that the critiques of the former are almost never heard in the latter. There is some sense to this, in that mathematics education at the grade school level involves every student, and by the time one arrives at university one already has entrenched preconceptions on who should excel at mathematics and who should not, and what mathematics is good for, and so on.</p>
<p>So let’s listen a bit on the conversation that has taken place in mathematics education: Alan Bishop has written on western mathematics as the secret weapon of cultural imperialism (1990), where western mathematics is understood to be itself a problematic term, having received essential contributions from Arab and Chinese mathematics, for example. Bishop writes that the process of cultural invasion by western mathematics in colonised countries has three major mediating agents: trade, administration, and education. The first two are quite routine, so I’ll focus on the third, also because it resonates with the theme that abstract mathematics is in fact political. The mathematics curriculum is embedded in the education of local elites, producing disciplined subjects in a European mould, that is, in the image of the coloniser. Indeed, as university-preparatory education, students were trained to aspire towards attending western universities, and the aspirations of the students were towards attending western universities, ‘educated away from their culture and away from their society.’</p>
<h4 id="towards-anti-racist-mathematics">Towards anti-racist mathematics</h4>
<p>Just as we have seen US military tactics abroad eventually come home to roost, so is the use of mathematics as a tool of subjugation in colonies brought to bear on a nation’s own citizens. Embedded within the methodology of western mathematics are the cultural values of rationalism (reasoning and logic), objectism (decontextualising in order to generalise), control (power over physical and social environments), and progress (industrialisation and development). These stand in stark contrast to the assumption that mathematics is universal and culturally neutral. Indeed, critical scholars have argued that mathematics education today is increasingly influenced by neoliberal and neoconservative market-focused projects, as we have already seen in military development and the commodification of students. Such projects can be seen as consequences of the cultural values embedded in western mathematics.</p>
<p>Arguably on the fringes of mathematics education (and non-existent in modern mathematics research) is the field of <a href="https://en.wikipedia.org/wiki/Ethnomathematics">ehtnomathematics</a>. While perhaps noble in its intention to explore and acknowledge the mathematical knowledge of nonwestern cultures (or even western cultures, in the case of indigenous americans), the project of ethnomathematics is fraught with the risk of simply reproducing the racial categories and hierarchies, becoming co-opted into ‘diversity’ efforts.</p>
<p>Martin identifies mathematics education itself as a <a href="http://education.uic.edu/sites/default/files/Race_Racial_Projects.pdf">racial project</a>, that is, ‘simultaneously an interpretation, representation, or explanation of racial dynamics and an effort to reorganize or redistribute resources along particular racial lines.’ For example, we see this in the distribution of scores among US students in standardised tests like the SAT. This makes plain the imperative to locate whiteness in mathematics as the gateway to higher education and access, and anti-racist mathematics as an answer to it.</p>This is an expanded version of the general audience talk I’m giving at the Science on Tap–formerly known as Drunk on Science, which is much cooler but probably less professional–in Pune, the brainchild of the illustrious Anoop Mahajan, with craft beer sponsored by Great State Ale Works! See Part 1, Part 2, Part 3, Part 4How to use math for evil III: Cryptosystems2017-10-03T11:50:45+00:002017-10-03T11:50:45+00:00https://tiwong.github.io/data/2017/10/03/how-to-use-math-for-evil-3<p><em>This is an expanded version of the general audience talk I’m giving at the Science on Tap–formerly known as Drunk on Science, which is much cooler but probably less professional–in Pune, the brainchild of the illustrious Anoop Mahajan, with craft beer sponsored by Great State Ale Works! See <a href="https://tiwong.github.io/data/2017/09/28/how-to-use-math-for-evil-1.html">Part 1</a>, <a href="https://tiwong.github.io/data/2017/09/30/how-to-use-math-for-evil-2.html">Part 2</a>, <a href="https://tiwong.github.io/data/2017/10/03/how-to-use-math-for-evil-3.html">Part 3</a>, <a href="https://tiwong.github.io/data/2017/10/06/how-to-use-math-for-evil-4.html">Part 4</a></em></p>
<p>Last time we entered into world of Big Data, and saw how complicated algorithms extend into our daily lives. We ended on the fact that Big Data is contingent upon data mining, which drills deeper and deeper into our lives. This time, we’ll talk about surveillance and cryptosystems, that are intertwined with this process.</p>
<h4 id="death-by-algorithm">Death by algorithm</h4>
<p>Before we get started, let’s note that mathematics has been used in warfare from ages past. Trigonometry and calculus have helped to calculate trajectories of rockets and bullets, a study known as <a href="https://en.wikipedia.org/wiki/Ballistics">ballistics</a>. If you watch shows like CSI, this is also involved in forensics, like with blood-splatter analysis.</p>
<p>And in keeping with the theme from Part 2, Obama’s so-called presidential kill list, formally known as the <a href="https://en.wikipedia.org/wiki/Disposition_Matrix">Disposition Matrix</a>, that governs drone strikes in the Middle East and Africa in fact involves complicated algorithms like the NSA’s <a href="https://en.wikipedia.org/wiki/SKYNET_(surveillance_program)">Skynet program</a> (not the the one in Terminator!) applies machine learning to communications data, in order to identify possible terror suspects. A terror suspect is then assigned certain scores like a player on a baseball card, and this ranks them in the kill list, which in turn determines drone strikes. <a href="https://www.theguardian.com/commentisfree/2016/feb/21/death-from-above-nia-csa-skynet-algorithm-drones-pakistan">In the words</a> of former CIA and NSA director General Michael Hayden himself,</p>
<blockquote>
<p>“We kill people based on metadata.”</p>
</blockquote>
<p>For one, The person targeted by the drone strike is merely a <em>suspect</em>, and we know this to be faulty knowledge at times from various reports over the years. Secondly, any civilian casualties, of which there are many, gets written off as collateral damage.</p>
<p>He later qualified his statement that the US government only kills foreigners, not US citizens this way. But even if you were a US citizen, you certainly have reason to worry. As mentioned in the last post, US warfare tactics abroad do find their way back home. Good examples are the counterinsurgency measures used at police-brutality protests like <a href="https://gizmodo.com/welcome-to-the-new-age-of-counterinsurgency-policing-1702621152">in Ferguson and Baltimore</a>, and counterterrorism tactics used at the <a href="https://theintercept.com/2017/05/27/leaked-documents-reveal-security-firms-counterterrorism-tactics-at-standing-rock-to-defeat-pipeline-insurgencies/">Dakota Access Pipeline protest</a> at Standing Rock. Not to mention the fact that the militarisation of US police force is a direct outcome of government policy like the 1997 <a href="https://en.wikipedia.org/wiki/1033_program">1033 program</a> and <a href="https://www.theguardian.com/us-news/2017/aug/28/donald-trump-news-police-military-style-weapons-vehicles">more recent presidential orders</a>, giving surplus military equipment to civilian law enforcement services.</p>
<p>Also of note is the possibility that torture is being optimised for effectiveness by <a href="https://www.nytimes.com/2017/08/12/opinion/sunday/when-torture-becomes-science.html">collecting data on methodology</a>. This is simply a logical extension of the data-driven approach to anything. A more innocuous example is how gerrymandering is also <a href="https://www.nytimes.com/2017/10/06/opinion/sunday/computers-gerrymandering-wisconsin.html">becoming a science</a>, by simulating all possible voting scenarios by a clever use of algorithms. Brining the math to bear on politics is an uphill battle, as it seems that the US Supreme Court is <a href="https://fivethirtyeight.com/features/the-supreme-court-is-allergic-to-math/amp/">allergic to math</a>.</p>
<h4 id="public-key-cryptography">Public-key cryptography</h4>
<p>The advent of quantum mechanics, what one might think of as <a href="https://en.wikipedia.org/wiki/Blue_skies_research">blue-sky research</a>, that is, research with no immediate ‘real-world’ applications, also came hand in hand in nuclear fission, as with the infamous <a href="https://en.wikipedia.org/wiki/Manhattan_Project">Manhattan Project</a>. The same has happened for number theory, which the famous mathematician Gauss called the Queen of Mathematics. Number theory is the study of prime numbers, the building blocks of integers. It is also the foundation for modern cryptosystems, or encryption methods. They do good things, like protect your credit card purchase on Amazon from being hacked and keep your bank accounts secure, but they also are at the epicentre of issues surrounding surveillance, security, and privacy.</p>
<p>The basic idea of encryption is simple: Alice wants to tell Bob a secret, and if Eve gets a hold of the message, the message would read like gibberish. To do this, Ann must encrypt the message in a way that only Bob can decipher. Here’s how it goes: Alice starts with an algorithm that enciphers the message into numbers, say the way you might have done as a school girl:</p>
<div class="language-plaintext highlighter-rouge"><div class="highlight"><pre class="highlight"><code>I C E C R E A M ---> 9 3 5 3 18 5 1 13
</code></pre></div></div>
<p>Now the trick is to use a cipher key, which let’s say is ‘multiply by 3’, so that you get
<code class="language-plaintext highlighter-rouge">27 9 15 9 54 15 3 39</code>, and then Bob, who also has the key can ‘unlock’ the encrypted message. The trouble with this system is that Eve can intercept this, and if she figures out the key, learns the secret. In this method the keys have to constantly be changed, and it’s a complicated process for Alice to tell Bob the new key, because if she uses the old key and Eve knows the old key, then Eve also gets the new key. But this was the way things worked for a long time like the <a href="https://en.wikipedia.org/wiki/Enigma_machine">Enigma Machine</a> in World War II, and being able to crack secrets was highly in demand during times of war.</p>
<p>Then in 1976 came the <a href="https://en.wikipedia.org/wiki/Diffie–Hellman_key_exchange">Diffie-Hellman key exchange</a>, one of the first public-key protocols. This introduced the asymmetric cipher, where the encrypting and decrypting keys are different, so Alice and Bob can now share a secret even if Eve is listening, since Alice does not have to communicate any key to Bob.</p>
<p>The <a href="https://en.wikipedia.org/wiki/RSA_(algorithm)">RSA encryption scheme</a>, invented the next year, put this into practice. And this is where number theory comes in: the fact that the procedure works is guaranteed by <a href="https://en.wikipedia.org/wiki/Fermat%27s_little_theorem">Fermat’s Little Theorem</a>. Here’s a rundown of how it works:</p>
<ol>
<li>Alice picks two prime numbers, let say <code class="language-plaintext highlighter-rouge">p</code> and <code class="language-plaintext highlighter-rouge">q</code>. Their product <code class="language-plaintext highlighter-rouge">N = pq</code> is the public key.</li>
<li>Alice picks a number <code class="language-plaintext highlighter-rouge">e</code> relatively prime to <code class="language-plaintext highlighter-rouge">p-1</code> and <code class="language-plaintext highlighter-rouge">q-1</code>. This is also made public.</li>
<li>Bob wants to tell Alice a message <code class="language-plaintext highlighter-rouge">M</code>, so he encrypts it as <code class="language-plaintext highlighter-rouge">C = M^e (mod N)</code> and sends this.</li>
<li>Alice solves for a number <code class="language-plaintext highlighter-rouge">d</code> such that <code class="language-plaintext highlighter-rouge">ed = 1 (mod (p-1)(q-1))</code>.</li>
<li>Alice decodes by calculating <code class="language-plaintext highlighter-rouge">C^d (mod N)</code>, and the answer is equal to <code class="language-plaintext highlighter-rouge">M</code>.</li>
</ol>
<p>Here, <code class="language-plaintext highlighter-rouge">mod</code> stands for modulo, as in <a href="https://en.wikipedia.org/wiki/Modular_arithmetic">modular arithmetic</a> or ‘clock’ arithmetic. It works like a clock: let’s start at 9 o’clock. Then any multiple of 12 hours, say 36, later is still 9 o’clock. That is, <code class="language-plaintext highlighter-rouge">45 = 9 (mod 12)</code>. In our example, to reduce <code class="language-plaintext highlighter-rouge">C^d (mod N)</code>, simply take out as many multiples of <code class="language-plaintext highlighter-rouge">N</code> from <code class="language-plaintext highlighter-rouge">C^d</code>, and what’s left over is your answer, which by some mathematics turns out to be <code class="language-plaintext highlighter-rouge">M</code>.</p>
<p>In real life, the primes are much larger, on the order of hundreds of digits long, so that factoring <code class="language-plaintext highlighter-rouge">N</code> would be practically impossible unless you were Alice or Bob. This is necessary since the decryption key requires knowing the values <code class="language-plaintext highlighter-rouge">p-1</code> and <code class="language-plaintext highlighter-rouge">q-1</code>. Also, in practice, public-key takes a huge amount of computer time, so what really happens when you tell Amazon your credit card number is a hybrid of public <em>and</em> private key cryptography.</p>
<p>As you can see, public-key cryptography depends on certain mathematical operations (like factoring!) being hard. More sophisticated methods have been developed, balancing security and efficiency as one often comes at the cost of the other. An important example comes from the world of number theory, called <a href="https://en.wikipedia.org/wiki/Elliptic-curve_cryptography">Elliptic Curve Crytography</a>, or ECC. It uses a certain multiplicative rule on an elliptic curve that looks like <code class="language-plaintext highlighter-rouge">y^3 = x^3 + ax + b</code>. There are even recommended elliptic curves, i.e., parameters <code class="language-plaintext highlighter-rouge">a</code> and <code class="language-plaintext highlighter-rouge">b</code> for government use.</p>
<h4 id="big-brother-and-big-data">Big brother and big data</h4>
<p>Now what’s important to the NSA is that such recommended cryptosystems have backdoors, which are basically cheat codes in this secret sharing game that allow the messages to be intercepted and decrypted at will. That is to say, your data may be secure to most people, but to the professionals, shall we call it, your data is not. The cold, hard logic behind this is always under the banner of national security.</p>
<p>And it is for this reason that your data, that is being mined by large corporations for profit, is also open access to the state, at least if said corporations are US firms. Indeed, among Edward Snowden’s leaks is that tech companies like Microsoft, Google, Facebook, and Apple have <a href="https://www.theatlantic.com/technology/archive/2014/07/the-details-about-the-cias-deal-with-amazon/374632/">tapped their own servers</a> for the intelligence agencies such as the NSA and FBI, while Amazon Web Services is a <a href="https://www.theatlantic.com/technology/archive/2014/07/the-details-about-the-cias-deal-with-amazon/374632/">CIA contractor</a>.</p>
<p>Perhaps most baffling of all is Signal, the go-to encrypted chat app used by US journalists and activists, endorsed even by Snowden. It is also adopted into Facebook’s Whatsapp. The irony is that Signal’s parent company <a href="https://en.wikipedia.org/wiki/Open_Whisper_Systems">Open Whispers Systems</a> has received almost 3 million in US government funding to date via Radio Free Asia. There are reasons for this, discussed along with a more secure chat app, Telegram, and its Russian inventor and dissident Pavel Durov in <a href="https://thebaffler.com/salvos/the-crypto-keepers-levine">this exposée</a>. Certainly the CIA has taken much more indirect routes, as with abstract expressionism being used as a <a href="http://www.bbc.com/culture/story/20161004-was-modern-art-a-weapon-of-the-cia">Cold War weapon</a>.</p>
<p>This makes the deeper data mining of companies more unsettling: the increasing amounts of surveillance performed by tech giants for profit, can easily be obtained—whether by direct requests or third-party hacking—by US intelligence agencies for security. The recent requirement of the Department of Homeland Security, or DHS, to <a href="https://www.nytimes.com/2017/09/28/us/politics/immigrants-social-media-trump.html">collect social media information</a> on immigrants to the USA is one such frontier. The trouble with ‘security’ is that it is not possible to define or set fixed boundaries for, and we see its full power in more totalitarian regimes where there is no façade of democracy. Also, a recent <a href="http://openscholarship.wustl.edu/cgi/viewcontent.cgi?article=6265&context=law_lawreview">journal article</a> makes the case that surveillance and data collection burdens poor people ‘many times over,’ consistent with the topic of Part 2.</p>
<p>Before, Michel Foucault’s <a href="https://en.wikipedia.org/wiki/Panopticism">Panopticon</a> as a symbol of the surveillance used to discipline society’s members by making each person the ‘principle of his own subjection’ has now grown into words like ‘dataveillance’ and ‘information panopticon.’ But the difference now, with the advent of Big Data, is that we can be sure we are being watched. The slow march towards complete surveillance removes the need for a Panopticon.</p>This is an expanded version of the general audience talk I’m giving at the Science on Tap–formerly known as Drunk on Science, which is much cooler but probably less professional–in Pune, the brainchild of the illustrious Anoop Mahajan, with craft beer sponsored by Great State Ale Works! See Part 1, Part 2, Part 3, Part 4How to use math for evil II: Algorithms2017-09-30T11:50:45+00:002017-09-30T11:50:45+00:00https://tiwong.github.io/data/2017/09/30/how-to-use-math-for-evil-2<p><em>This is an expanded version of the general audience talk I’m giving at the Science on Tap–formerly known as Drunk on Science, which is much cooler but probably less professional–in Pune, the brainchild of the illustrious Anoop Mahajan, with craft beer sponsored by Great State Ale Works! See <a href="https://tiwong.github.io/data/2017/09/28/how-to-use-math-for-evil-1.html">Part 1</a>, <a href="https://tiwong.github.io/data/2017/09/30/how-to-use-math-for-evil-2.html">Part 2</a>, <a href="https://tiwong.github.io/data/2017/10/03/how-to-use-math-for-evil-3.html">Part 3</a>, <a href="https://tiwong.github.io/data/2017/10/06/how-to-use-math-for-evil-4.html">Part 4</a></em></p>
<p>So last time we looked at how statistics can be used and misused to advance certain agendas. That’s ancient history. This time we’ll turn our focus to the age of the smart phone: algorithms, and the murky waters of Big Data that they now swim in.</p>
<h4 id="models-and-machines">Models and machines</h4>
<p>What is an algorithm? At the most basic level, an algorithm is really just a set of instructions that a computer can follow. The bottomline of an algorithm is a bunch of computer code, but the algorithm describes a model, and the model is where it all begins. To emphasise the matter, there is a growing push towards <a href="https://www.theatlantic.com/technology/archive/2017/09/saving-the-world-from-code/540393/">model-based design</a> as an improvement of current programming practices, where focus is shifted from writing code that works to producing better models.</p>
<blockquote>
<p>“Models are opinions embedded in mathematics.”</p>
</blockquote>
<p>Let’s back track a little, and talk about what a model is. Suppose you want to quantify a certain process, say what are the chances someone surfing the web might click on a particular ad. You’d start with a person <code class="language-plaintext highlighter-rouge">x</code> on a webpage like Facebook. Then based on certain traits measured by say, their web activity and personal information, you’d figure out which kinds of ads they would respond to the best. For example, if you happened to like a lot of pictures with dogs in it, then you might get served more dog ads.</p>
<p>But even this innocuous scenario raises some points about technology that we will look deeper into later: Your user profile and activity are commodities, and are sold as such with Facebook, Google, and Amazon topping the list of data brokers. Of course, as person <code class="language-plaintext highlighter-rouge">x</code>, your data is supposedly anonymised so as to protect your privacy, but a general principle in mathematics is that if you know enough properties of an unidentified object, you might in fact be able identify said object uniquely. For example, just knowing one’s zip code, birthdate, and sex is enough to identify 87 percent of US citizens (<a href="https://arstechnica.com/tech-policy/2009/09/your-secrets-live-online-in-databases-of-ruin/">link</a>). But more on this later.</p>
<p>A second point touched on is the ability of a computer to recognise dogs. This is machine learning. It’s the idea behind self-driving cars and AI-composed music, which is that given enough historical data, you can teach a computer to generate its own new data. This is what’s important with breakthroughs and the buzzword <em>deep learning</em>. Getting a computer to <a href="https://www.scientificamerican.com/article/how-the-computer-beat-the-go-master/">master Go</a> is more sensational than chess is because a computer can easily store all possible chess moves, but there’s way too many possibilities in Go, so the computer has to be able to adapt to its opponents specific plays, not having a list of all possible scenarios in hand. Kind of like teaching a car to drive.</p>
<p>Of course, this has led to certain faux pas like Google’s recognition software tagging <a href="http://www.wnyc.org/story/deep-problem-deep-learning/">black people as gorillas</a> and Microsoft’s chatbot <a href="https://www.theverge.com/2016/3/24/11297050/tay-microsoft-chatbot-racist">turning Nazi</a> by learning from Twitter in a day. These are slightly offensive and slightly funny scenarios, but the basic fact is that AI learns from human behaviour (in the form of data), and in the process it learns our biases and bigotries. Interestingly, this is sort of like an automated form of discourse analysis as in Edward Said’s conclusion of orientalism of various Western literary works; one could learn to be racist with Twitter as a textbook.</p>
<p><em>During the talk I presented some card tricks as demonstrations of algorithms. <a href="http://www.cs4fn.org/magic/">Here’s my source</a>. Check it out!</em></p>
<h4 id="big-data-in-black-boxes">Big data in black boxes</h4>
<p>Now let’s scale up a little, and look at certain algorithms that Cathy O’Neil calls Weapons of Math Destruction, or WMDs for short. This comes from her book <em>Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy</em>. In it, she defines a WMD to be an algorithm that has: (1) opacity, (2) scale, and (3) damage. Let’s break these down a little.</p>
<p>An algorithm is opaque when its processes or its inner logic, i.e., the model that it operates on, is invisible to the people it is being applied to. For example, most of us don’t know how ad services work exactly, though we might have a few guesses. We call this a black box. It’s a term in used in computer science and mathematics, to describe a tool which you don’t know exactly why or how it works, but you use it anyway, taking for granted that it does because someone else vetted it. Most algorithms that operate in the background of our lives are opaque, under some kind of patent or corporate lock. Importantly, the point is not that everyone can understand it, but if a third party want to audit the code for say, certain biases, they would have access to do so.</p>
<p>Scale is a simple enough feature: the algorithm must be rolled out at a large scale, so that it is affecting many people, and hence has many potential victims. Which is the third point: damage. The algorithm must be affecting people negatively, and in a way that discriminates against people with particular traits, like race and income. When these three come together, O’Neil calls the algorithm a Weapon of Math Destruction. It has other features, like so-called ‘pernicious feedback loops’. That is, the algorithm’s success is measured by the results that it gets, and if its results reinforce an inequality, that reproduces more bad data like the kind that went into the model that it was built on in the first place.</p>
<p>Of course, the algorithm’s black box is precisely what gives it the impression of impartiality and veracity: it is a machine, so it should be less prone to emotional judgments and prejudices than a human (although the machine itself learned from humans), and that its results are God’s truth, since we can’t see through the black box, and even if we could we probably wouldn’t understand that model or the mathematics behind it all anyway. As they say, the numbers don’t lie. Or do they?</p>
<h4 id="some-wmds-in-the-wild">Some WMDs in the wild</h4>
<p>Here’s a quick and dirty run through of most of the examples covered in the book. The book is written by a US author for a US audience, so the case studies will be very much focused on US society and politics.</p>
<ol>
<li>
<p><strong>College rankings</strong>. In 1983, a group of journalists at the US News & World Report came up with a list of measurable criteria, i.e., a model, for what makes a college or university the ‘best’. This set of an arms race among schools to boost performance in these criteria as the rankings became the national standard. One inherent problem is that this measuring stick is not a one-size-fits-all standard: what’s best for you is not necessarily what’s best for me. Just as this is true in terms of nutrition, so it is in education. At the same time, this sets up a system that can be gamed. For example, certain law schools boosting post-graduation employment rates by giving their own graduates temp jobs.</p>
<p>The second problem is a simple pernicious feedback-loop: How do the modellers know that their model is accurate? In this case, the rankings would be more credible if they reflect the pre-existing hierarchy. Elite schools, i.e., schools in which children of the elite enrolled, like Harvard and Princeton had to top the list to validate the model. In other words, the rankings perpetuate the existing inequalities and reinforce class lines. Importantly, an important factor that is left out of the rankings is the tuition price. This effectively gives schools permission to raise tuition to generate more revenue to boost performance in the ranking’s metrics (which has happened nationwide).</p>
<blockquote>
<p>If you look at this development from the perspective of a university president, it’s actually quite sad. Most of these people no doubt cherished their own college experience—that’s part of what motivated them to climb the academic ladder. Yet here they were at the summit of their careers dedicating enormous energy toward boosting performance in fifteen areas defined by a group of journalists at a second-tier newsmagazine. They were almost like students again, angling for good grades from a taskmaster. In fact, they were trapped by a rigid model, a WMD. – O’Neil</p>
</blockquote>
<p>Of course, the victims here end up being folks who cannot afford the sticker price on the one hand, being led into debt, and on the other hand, cannot afford tutors and counsellors who will coach students through the admissions process. Each college has an admissions model tailored to the maximise the college’s own rankings, and are mini WMDs in themselves, driving the race for SAT scores and extraordinary extracurriculars.</p>
</li>
<li>
<p><strong>Recidivism risk</strong>. Models like LSI-R (Level of Service Inventory—Revised) and <a href="https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm">COMPAS </a> are tools used by judges to help determine the severity of sentences handed out. Using certain information including a questionnaire that is filled out by the guilty person, it determines the risk of them committing a crime again. This makes sense at first, taking the possible moods or biases of the judge out of the equation. But the questions collect historical data that is already biased. Questions like: How many previous convictions have you had, when was the first time you were involved with the police, and Do you know anyone who has been convicted of a crime, are questions whose answers depend more on the economic and racial demographics of a certain neighbourhood. While not legally allowed to ask one’s race, and seemingly innocent information like one’s zip code, as discussed above, become proxies for race and class. Indeed, recidivism risk combines with the next model as a part of a big negative feedback loop:</p>
<p><strong>Predictive policing</strong>. Very often, algorithms are meant to optimise a certain process, given limited resources. In this case, the optimisation problem involved is how to distribute police patrolling in areas that where crime is more likely to occur. In New York City, for example, this is carried out on the ground in the form of policies like <a href="https://en.wikipedia.org/wiki/Broken_windows_theory">Broken Windows</a> and, more recently, <a href="https://en.wikipedia.org/wiki/Stop-and-frisk_in_New_York_City#Court_cases">Stop and Frisk</a>. But this is also implemented through algorithms like PredPol and Compstat, which predict neighbourhoods that are hot spots for criminal activity. In this case, it’s quite easy to see how the pernicious feedback loop is generated: more policing leads to more arrests, and more arrests lead to more policing. It’s important to note that zero-tolerance policies like Broken Windows causes police to not distinguish between so-called Part 1 crimes (homicide, assault, etc.) and Part 2 crimes (aggressive panhandling, selling or consuming small quantities of drugs, etc.). Part 1 crimes are crimes that should be prevented without question, while Part 2 crimes are often situational; think of soft drugs being used on a college campus versus a ‘bad’ neighbourhood.</p>
<p>And of course, this model only picks up ‘blue-collar’ crimes: acts such as fraud, embezzlement, and economic catastrophes like the 2009 financial crisis slip by silently. As such, increased police activity on the ground makes living in certain zip codes more of a liability, and you can imagine how that goes into a recidivism risk calculation. Also worth mentioning is the fact that counter-terrorism and counter-insurgency practices of the US army abroad (which we’ll cover in the next post!) are slowly brought home to roost. In this case, as a look into the future, facial recognition software is being tested to identify not just wanted persons, but also potential criminals. Fortunately, this was turned down by the San Diego police department in favour of privacy. Small victory.</p>
</li>
<li>
<p><strong>Human Resources</strong>. We’ll look first at getting a job, and then what happens on the job. As you can imagine, getting a job with the deck stacked against you like before is easier for some more than others. When you put in a job application, your application or interview might include questions that are basically a psychological profile or personality test, like those developed by <a href="https://en.wikipedia.org/wiki/Kronos_Incorporated#Litigation_with_EEOC">Kronos</a>, a workforce management company, based on the <a href="https://en.wikipedia.org/wiki/Big_Five_personality_traits">Five Factor Model</a>. This sounds innocuous, but choosing between “It is difficult to be cheerful when there are many problems to take care of” or “Sometimes, I need a push to get started on my work” might be the difference between being identified as “conscientiousness” versus “low ambition and drive”. This profiling says nothing certain about a candidate’s ability to perform their job, but now becomes a deciding factor on getting a job. And, as you can guess, those with any mild personality disorders or mental disabilities get the short end of the stick.</p>
<p>Algorithms like these are mostly used by large companies to streamline their hiring process. Even with a smaller company, your resume might be read by a computer that sorts through thousands of others, and it’s your job to know the keywords and phrases that will make sure your resume survives this initial culling process. And knowing what a winning resume looks like, again takes resources more readily available to the more privileged. Of course, hiring practices have been discriminatory long before the help of automation; it’s simply that the biases get built into the algorithms, and become opaque. Also, the use of these systems is not uniform: it gets used less and less as you go up the ladder, and most on those entry level, low-wage jobs. It’s quite likely that such models not only propagates the existing power structure, they elaborate upon it, streamline it.</p>
</li>
<li>
<p><strong>Operations research</strong>. Let’s suppose that after all that, you managed to land a job. As a worker bee, you’ve in fact entered as worker <code class="language-plaintext highlighter-rouge">x</code> into the world of operations research, or OR. It is properly a branch of applied mathematics, meant to optimise efficiency in growing crops and moving goods. This was also used in <a href="https://en.wikipedia.org/wiki/Operations_research#Historical_origins">World War II</a>, to minimise the number of mine-laying aircraft for each Japanese merchant ship delivering supplies to Japanese shores. In the 1960s, Japan developed a manufacturing system called <a href="https://en.wikipedia.org/wiki/Just-in-time_manufacturing">Just in Time</a> that minimises the time that parts in delivery remained idle, increasing efficiency and saving cost across the supply chain.</p>
<p>OR is now much more sophisticated, micromanaging the worker down to the minute. This sometimes goes under the banner of management science. For example, companies like Starbucks use scheduling software that analyses customer traffic to schedule its employees, and Uber even <a href="https://www.nytimes.com/interactive/2017/04/02/technology/uber-drivers-psychological-tricks.html">conducts behavioural science experiments</a> on its drivers to optimise its labor force in a gig economy. Unoptimised, this analysis is quite harmless, and makes simple recommendations; but when optimised to be sensitive to mercurial variables such as weather and Twitter volume, the employee’s work hours are held hostage by a machine.</p>
<blockquote>
<p>“The trouble, from the employees’ perspective, is an oversupply of low-wage labor. People are hungry for work, which is why so many of them cling to jobs that pay barely eight dollars per hour. This oversupply, along with the scarcity of effective unions, leaves workers with practically no bargaining power. This means the big retailers and restaurants can twist the workers’ lives to ever-more-absurd schedules without suffering from excessive churn. They make more money while their workers’ lives grow hellish. And because these optimization programs are everywhere, the workers know all too well that changing jobs isn’t likely to improve their lot. Taken together, these dynamics provide corporations with something close to a captive workforce.”</p>
</blockquote>
<p>Another instance of OR is the so-called <a href="https://en.wikipedia.org/wiki/Value-added_modeling">value-added school teacher performance measure</a>, which is basically a lesson in statistics from the last post. In 1983 the Reagan administration released a report called <em>A Nation at Risk</em>, warning of a “rising tide of mediocrity”, exemplified by falling SAT scores. As it turns out, it was a case of <a href="https://en.wikipedia.org/wiki/Simpson%27s_paradox">Simpson’s Paradox</a>: but many more students were taking the SAT, which brought down the total average, but when broken down into income groups, every single one of the individual averages had risen. Nonetheless, alarms such as these has led to assessing teachers with numbers. The value-added model tries to figure out how much of a student’s outcome can be attributed to a teacher’s ability. A simple but major problem is that this model works if the teacher taught millions of students, and all the students’ outcomes could be analysed and assessed accurately. But in reality a teacher teaches only a handful of students, and making any statistical prediction an educated guess, at best.</p>
</li>
<li>
<p><strong>Predatory advertising</strong>. Now let <code class="language-plaintext highlighter-rouge">x</code> be a consumer. Our next WMD in line is engineered to optimise the amount of dollars extracted from <code class="language-plaintext highlighter-rouge">x</code>, given a set of data associated to <code class="language-plaintext highlighter-rouge">x</code>. Okay, that sounds a little to math-y, but that’s really how these black boxes are built: the machination of capitalist processes requires converting flesh and blood into numbers and data points. But as we’ve already seen, only crude models don’t see the individual. Sophisticated models learn from your data to adapt to your specific profile, and fine tune advertising campaigns to appeal to <em>you</em>.</p>
<p>O’Neil makes this transition by discussing for-profit colleges, who use predatory advertising techniques to seek out vulnerable, paying customers. Of course, these techniques existed before Big Data, like tobacco companies’ advertising campaigns. The basic idea is the same: to look for specific <em>pain points</em>, places where a person is suffering and is most vulnerable, then press on those points so the person buys the product, an impulse decision more emotional than rational. And if the person cannot afford to buy, offers pour in for high-interest payday loans. Of course, shady advertising is nothing new, but now an army of data scientists are recruited to optimise the targeting.</p>
</li>
<li>
<p><strong>Consumer profiles</strong>. One method of online targeting is known as <a href="https://en.wikipedia.org/wiki/Lead_generation">lead generation</a>, which are ways to obtain the data that will generate a list of prospects to target. We’ll skip the more shady techniques to the big time data brokers: it goes without saying that the questions we ask Google range from innocuous to truly revealing. Actually, you don’t even have to ask. The information extracted from your activity on all of Google’s products or Facebook’s apps—which include <a href="https://en.wikipedia.org/wiki/List_of_mergers_and_acquisitions_by_Facebook">Whatsapp and Instagram</a>—is sold as data to any paying customer, including these predatory advertising agencies.</p>
<p>The data scientist starts with simple Bayesian statistics, ranking different advertising techniques by the probability that they are effective, i.e., translate to the most dollars. Then she proceeds to run a series of competing ads against each other to test for effectiveness. This is known as <a href="https://en.wikipedia.org/wiki/A/B_testing">A/B testing</a>, used by direct-mail marketers decades ago. But with online advertising to web traffic as large as those on Google or Facebook, this testing becomes extremely effective (as was the opposite with teacher evaluations) and individualised.</p>
<p>Consumer profiles are created as ‘buckets’, which is short for groups of people just like you. It often consists of data such as your zip code and e-score. Your zip code, as we have seen, is a stand in for race and class. Your e-score, is an unregulated proxy for your credit score, the latter being illegal to use for marketing purposes. The e-score is hammered together out of your data such as geo-tags and clickstreams. In more nefarious situations, e-scores have been used to assess candidates for job offers. Also, analytics obtained from smarter and smarter cars get sold to insurance agencies to reduce the payouts they have to make on claims. The irony is that it is the job of risk managers (and actuarial scientists) to reduce this spending, whereas the whole purpose of insurance is for people be financially protected from the risk of unlikely, unwanted costs.</p>
</li>
<li>
<p><strong>Voter profiles</strong>. Just as your behavioural data is fodder for private companies, so have political campaigns used this data to serve up advertisements tailored to specific voter demographics. While the efficacy of these are debated, the most sophisticated tools were certainly used in the last 2016 US presidential election. Specifically, targeted marketing, or <a href="https://en.wikipedia.org/wiki/Microtargeting">microtargeting</a>, is brought to bear on swing voters the way a company would try to convince a potential customer to switch brands. This affects the nature of democracy, putting the fate of an election in the hands of a select few (and of course, the US electoral college).</p>
<p>Facebook’s newsfeed algorithm (nevermind the fake news) serving up different politics to different groups was well-documented and discussed in the Wall Street Journal’s <a href="http://graphics.wsj.com/blue-feed-red-feed/">Red Feed, Blue Feed</a> experiment. It’s a textbook case of machine learning and A/B testing creating digital silos, or echo chambers. Also worth noting is Facebook’s ‘I voted!’ badge and its <a href="http://www.motherjones.com/politics/2014/10/can-voting-facebook-button-improve-voter-turnout/">influence on voter participation</a>, the results of the effect in 2010 were measured and published by a team of academics and Facebook data scientists in the prestigious Nature journal, under the title <a href="http://fowler.ucsd.edu/massive_turnout.pdf">A 61-Million-Person Experiment in Social Influence and Political Mobilization</a>. Is it acceptable for Facebook to quietly perform social experiments on its users, directly influencing politics? That’s up for debate. (A subsequent <a href="https://www.theatlantic.com/technology/archive/2014/06/everything-we-know-about-facebooks-secret-mood-manipulation-experiment/373648/">2013 experiment</a> showing that manipulating the Facebook News Feed in different ways could make users feel more positive or negative through ‘emotional contagion’ caused an outcry.)</p>
</li>
<li>
<p><strong>Quantitative finance</strong>. I’ll only cover this briefly, because the finance system is a real big behemoth, and it runs tirelessly in the backdrop of the global economy, driving the capitalist engine relentlessly forward. The idea of quantitative finance, is that financial markets can be modelled like a physicist might model air flow, as with <a href="https://en.wikipedia.org/wiki/Brownian_model_of_financial_markets">Brownian motion</a>. This has led to a surge of computer scientists, physicists, and even linguists into the trading floor. Financial mathematics is now a fully-fledged field of academic field drawing from pure mathematics like probability, like <a href="https://en.wikipedia.org/wiki/Monte_Carlo_method">Monte-Carlo methods</a>, and partial differential equations, like the <a href="https://en.wikipedia.org/wiki/Black–Scholes_model">Black-Scholes model</a>. A trader’s worth is measured by a single number: the <a href="https://en.wikipedia.org/wiki/Sharpe_ratio">Sharpe ratio</a>, which is basically the profits made divided by the risks taken. This incentives the trader to take bigger risks to get bigger profits, which is really what happens in casinos, but a trader bets with someone else’s savings or pension.</p>
<p>The role of algorithms in the 2008 financial crisis was simply this: the bubble created by sub-prime mortgage loans in mortgage-backed securities, i.e., credit default swaps and synthetic collateralised debt obligations, or CDOs—buzzwords you can read whole books or watch documentaries about—could not have been predicted by the risk models for the simple fact that such a thing <em>never happened before</em>. There was no historical data, no precedent. Moreover, the risk models were crunching out numbers that were in part meant to convince the buyer that the securities were not as risky as they really were, but the opacity of the risk ratings became a smokescreen to most in the financial industry itself.</p>
</li>
</ol>
<h4 id="using-data-for-good">Using data for good</h4>
<p>As a silver lining, O’Neil ends with a few examples of good models, models that are being used for good in the world:</p>
<ul>
<li>The possibility of auditing algorithms like the firm founded by O’Neil, <a href="http://www.oneilrisk.com">ORCAA</a>.</li>
<li>A slavery model by <a href="https://madeinafreeworld.com">Made in a Free World</a> which assess the risk that a product involves slavery in its supply chain.</li>
<li>A homeless recidivism model by the <a href="http://ctb.ku.edu/en/south-oakland-shelter-homelessness">South Oakland Shelter</a> in Michigan, which determines the risk of homeless persons becoming homeless again, and answering this risk with proper intervention.</li>
<li>A predictive model for households of child abuse by Florida’s <a href="https://gcn.com/articles/2014/08/04/predictive-analytics-child-welfare.aspx">Department of Children and Families</a>, identifying households where child abuse may occur, and acting accordingly.</li>
</ul>
<p>An important takeaway of O’Neil’s book is that Big Data is not necessarily evil. But neither is it morally neutral; data analytics serve up increasingly precise predictions based on the increasing amounts of data available, and the bottomline is what one chooses to do with the results, as the good models above show. Many of the bad algorithms described above can be used to fight injustice rather than perpetuate it. In O’Neil’s words,</p>
<blockquote>
<p>“Big Data processes codify the past. They do not invent the future. Doing that requires moral imagination, and that’s something only humans can provide. We have to explicitly embed better values into our algorithms, creating Big Data models that follow our ethical lead. Sometimes that will mean putting fairness ahead of profit.”</p>
</blockquote>
<h4 id="mining-and-spying">Mining and spying</h4>
<p>Let’s come back to where Big Data comes from. The advent of Big Data has not been precipitated so much by the computer or even the internet, but rather the smart phone. The algorithms like the ones described above are contingent upon having troves of data, whereas the access that smart phones provide slowly become indispensable parts of the modern life, and the apps that provide us with this access all collect data from us as a vast resource for various purposes, and our data is the price we pay for using free apps such as these. As a single user, your location, identity, and user habits are not valuable, but they become so as a part of a million other data points.</p>
<p>As it is, in comparing <a href="https://www.theguardian.com/world/2017/aug/23/silicon-valley-big-data-extraction-amazon-whole-foods-facebook">data mining to oil drilling</a>, the shallow reserves of data have and are being mined, and the next frontiers are to drill deeper for more. We are starting to see our data being mined not just online, but offline as our phones track our location, Siri and Cortana and Alexa <a href="https://www.digitaltrends.com/mobile/is-your-smartphone-listening-to-your-conversations/">listen to our conversations</a>, and <a href="https://www.theatlantic.com/business/archive/2017/06/why-amazon-bought-whole-foods/530652/">Amazon takes over Whole Foods</a>. The drive to dig for more data encroaches deeper and deeper into every aspect of our lives, and now you can imagine that this surveillance is not limited to corporations but the state, and they are not independent of each other.</p>
<p>That’s the next post.</p>This is an expanded version of the general audience talk I’m giving at the Science on Tap–formerly known as Drunk on Science, which is much cooler but probably less professional–in Pune, the brainchild of the illustrious Anoop Mahajan, with craft beer sponsored by Great State Ale Works! See Part 1, Part 2, Part 3, Part 4