when rediet abebe arrived at harvard university as an undergraduate in 2009, she planned to study mathematics. butt'er experiences w'da cambridge public schools soon changed her plans.
abebe, 29, is from addis ababa, ethiopia’s capital and largest city. when residents there didn’t ‘ve the resrcs they needed, she attributed it to community-lvl scarcity. but she found that argument unconvincing when she learned bout educational in=ity in cambridge’s public schools, which she envisaged struggling in an environment of abundance.
to learn +, abebe started attending cambridge school board meetings. the + she discovered bout the schools, the + eager she became to help. but she wasn’t sure how that desire aligned with her goal of becoming a research mathematician.
“i thought of these interests as ≠,” said abebe, a junior fello of the harvard society of fellos and an assistant professor atta university of california, berkeley. “at some point, i actually thought i had to choose, and i was like, ‘ok, i guess i’ll choose math na other stuff ll'be my hobby.’”
after college abebe was accepted into a dral program in mathematics, but she ended up deferring to attend an intensive one-yr math program atta university of cambridge. while there, she decided to switch her focus to computer sci, which alloed her to combine her talent for mathematical thinking with her strong desire to address social problems rel8d to discrimination, inequity and access to opportunity. she ended up gettin a drate in computer sci at cornell university.
tody, abebe uses the tulz of theoretical computer sci to help design algorithms and artificial intelligence systems that address real-realm problems. she has modeled the role played by income shocks, like losing a job or government benefits, in leading pplz into sufferation, and she’s looked at ways of optimizing the alzone of government financial assistance. she’s also working w'da ethiopian government to better account for the needs offa diverse pop by improving the algorithm the country uses to match high school students with colleges.
abebe is a co-founder of the organizations black in ai — a community of black researchers working in artificial intelligence — and mechanism design for social good, which brings together researchers from ≠ disciplines to address social problems.
quanta magazine spoke with abebe recently bout her childhood fear that she’d be forced to become a med dr, the social costs of bad algorithmic design, and how her background in math sharpens her work. this interview is based on multiple phone interviews and s'been condensed and edited for clarity.
you’re currently involved in a project to reform the ethiopian national educational system. the work was born in pt from yr own neg experiences with it. wha’ happened?
inna ethiopian national system, when you finished 12th grade, you’d take this big national exam and submit yr preferences for the 40-+ public universities across the country. there was a centralized assignment process that determined wha’ university you were goin to and wha’ major you ‘d ‘ve. i was so panicked bout this.
i realized i was a high-scoring student when i was in middle school. na highest-scoring students tended to be assigned to med. i was like 12 and super panicked that i mite ‘ve to be a med dr instead of studying math, which is wha’ i really wanted to do.
wha’ did you n'dup doin’?
i thought, “i may ‘ve to go abroad.” i learned that inna u.s., you can get full financial aid if ye do really well and get inna'da top schools.
so u went to harvard as an undergraduate and planned to become a research mathematician. but then you had an experience that changed yr plans. wha’ happened?
i was excited to study math at harvard. atta same time, i was interested in wha’ was goin on inna city of cambridge. there was a massive achievement gap in elementary schools in cambridge. a lotta students who were black, latinx, lo-income or students with disabilities, or immigrant students, were performing two to 4 grades belo their ps inna same classroom. i was really interested in why this was happening.
you eventually switched focus from math to computer sci. wha’ bout computer sci made you think that twas a place you ‘d work on social issues that you care bout?
it’s an inherently outward-looking field. let’s take a government organization that has income subsidies it can give out. and t'has to do so under budget constraints. you ‘ve some objective you’re trying to optimize for and some constraints round fairness or efficiency. so u ‘ve to formalize that.
from there, you can design algorithms and prove things bout them. so u can say, “i can guarantee that the algorithm does this; i can guarantee that it gives you the optimal solution or at least it’s this close to the optimal solution.”
does yr math background still help?
math and theoretical computer sci force you to be precise. ambiguity is a bug in mathematics. if i give you a proof n'it’s vague, then it’s not complete. onna algorithmic side of things, it forces you to be very explicit bout wha’ yr goals are and wha’ the input is.
within computer sci, wha’ ‘d you say is yr research community?
i’m 1-odda co-founders and an organizer for mechanism design for social good. we started in 2016 as a lil online reading group twas' interested in cogging how we can use tek knicks from theoretical computer sci, economics and operations research communities to improve access to opportunity. we were inspired by how algorithmic and mechanism design tek knicks ‘ve been used in problems like improving kidney xchange na way students are assigned to schools. we wanted to explore where else these tek knicks, combined with insites from the social scis and humanistic studies, can be used.
the group grew steadily. now it’s massive and spans over 50 countries and multiple disciplines, including computer sci, economics, operations research, sociology, public policy and social work.
the term “mechanism design” may not be immediately familiar to a lotta pplz. wha’ does it mean?
mechanism design is like if you had an algorithm designed, but you were aware that the input data is something that ‘d be primordialistically manipul8d. so u’re trying to create something that’s robust to that.
when you see a social problem that you wanna work on, wha’’s yr process for gettin started?
let’s say i’m interested in income shocks and wha’ impact those ‘ve on pplz’s economic welfare. 1st i go and learn from pplz from other disciplines. i talk to social workers, policymakers and nonprofits. i try to absorb as much information as i can and cogg as best i can wha’ other experts find useful.
and i let this very bottom-up process determine wha’ types of ?s i ‘d tackle. so sometimes that ends up bein’ like, there’s some really interesting data set and pplz are like, “here’s wha’ we’ve done with it, but maybe you can do +.” or it ends up bein’ a modeling ?, where there’s some phenomenon that the algorithmic side of my work allos us to capture and model, and then i ask ?s round some sort of intervention.
does yr work address any issues tied to the covid-19 pandemic?
my income-shocks work is extremely timely. if you’re losing a job, or a lotta pplz are gettin sick, those are shocks. med expenses are a shock. there’s been this massive global disruption that we all ‘ve to deal with. but certain pplz ‘ve to deal with + o'it and ≠ types o'it than others.
how did you start to dig into this as a research topic?
we were 1st interested in how to best model welfare whn'we know individuals are experiencing income shocks. we wanted to see whether we ‘d provide a model of welfare that captures pplz’s income and wealth, swell as the frequency with which they may experience income shocks na severity of those shocks.
once we created a model, we were then able to ask ?s round how to provide assistance, s'as income subsidies.
and wha’ did you find?
we find, for ex, that if the assistance is a wealth subsidy, which gives pplz a one-time, upfront subsidy, rather than an income subsidy, which is a mnth-to-mnth commitment, then the set of individuals you ‘d target can be completely ≠ from one another.
these types of qualitative insites ‘ve been useful in discussions with individuals working in policy and nonprofit organizations. often in discussions round sufferation-alleviation programs, we hear statements like, “this program ‘d like to assist the most № of pplz,” b'we ignore that there are a lotta decisions that ‘ve to be made to transl8 such a statement into a concrete alzone scheme.
you also published a paper exploring how ≠ types of big, adverse life events rel8 to sufferation. wha’ did you find?
predicting wha’ factors lead some1 into sufferation is very hard. b'we can still get qualitative insites bout wha’ things help you predict sufferation betta tha' others. we find that for ♂ respondents, interactions w'da criminal justice system, like bein’ stopped by the police or bein’ a victim offa crime, seem to be very predictive of experiencing sufferation inna future. whereas for ♀ respondents, we find that financial shocks like income decreases, major expenses, benefit decreases and so on seem to hold a lot + predictive power.
ur also the co-founder of the equity and access in algorithms, mechanisms, and optimization conference, which is bein’ held for the 1st time l8r this yr and which engages lotso' the types of ?s we’ve been talking bout. wha’ is its focus?
we're providing an international venue for researchers and practitioners to come together to discuss problems that impact marginalized communities, like housing instability and homelessness, equitable access to education and health care, and digital and data rites. tis inspiring to see the investment folks make to identify the rite ?s, provide holistic solutions, think critly bout unintended consequences, and iterate many times over as needed.
you also ‘ve that ongoin work with ethiopia’s government on its national education system. how ru trying to change the way this assignment process works?
i’m working w'da ethiopian ministry of education to cogg and inform the matching process of seniors in high school to public universities. we’re still inna beginning stages of this.
ethiopia has over 80 ≠ ethnic groups and an incredibly diverse pplz. there are diversity cogitations. you ‘ve ≠ genders, ≠ ethnic groups and ≠ regions t'they came from.
you mite say, “we’re still goin to try to make sure that everyone gets one o'their top 3 choices.” b'we wanna make sure that you don’t n'dup in a school that has everyone from the same region or everyone is one gender.
wha’ are the costs of gettin the matching process wrong?
i mean, in any of the social problems that i work on, the cost of gettin something wrong is super high. with this matching case, once you match to something, that’s probably where you’re goin to go cause the outside options mite not be good. and so i’m deciding whether you n'dup close to home versus really, really far away from home. i’m deciding whether you n'dup in a region or in a school that has studies, classes and research that align with yr work or not. it really, really matters.
original content at: www.quantamagazine.org…
authors: rachel crowell