Immediately after winning awards for technical innovations in financial theory and industrial organization, serving as main economist at Microsoft, conducting original investigate that brings together equipment finding out with econometric modeling, groundbreaking the new field of tech economics, helping to discovered and foster Stanford HAI, and launching the Golub Capital Social Effect Lab at Stanford, Susan Athey will now try on a new hat as chief economist at the antitrust division of the U.S. Division of Justice (DOJ).
The wide-ranging mother nature of her get the job done and career has garnered her regard throughout several tutorial fields. “Susan is a pressure of character. She moves from equipment discovering to enterprise method to know-how plan to social influence, producing deep tips at each individual switch,” states Jonathan Levin, the Philip H. Knight Professor and Dean at Stanford Graduate Faculty of Small business, wherever Athey is the endowed Economics of Know-how Professor.
Whilst Athey will carry on her GSB appointment in a part-time potential as she techniques into her new authorities role, the change in her target features an chance to reflect on the important impression she has had throughout her job and even though at Stanford.
A Role Design for HAI
Handful of researchers greater exemplify the multidisciplinary state of mind fostered by Stanford HAI than Athey. Her propensity for immersing herself in numerous fields of examine goes again to her undergraduate times at Duke University, the place she graduated in 1991 with a triple-major in economics, arithmetic, and laptop or computer science.
She went on to turn into a professor of economics and small business at MIT, Harvard College, and then at Stanford starting off in 2013, but even within just the discipline of economics, Athey’s pursuits have been diverse: In 2007, she gained the prestigious John Bates Clark Medal for her contributions to a number of subfields, including industrial organization, microeconomic theory, and econometrics.
But it was in the course of a go away from academia to serve as Microsoft’s chief economist from 2008 to 2013 that Athey produced a astonishing link between her passion for economics and the instruments of AI and machine discovering.
Athey currently realized that digitization and tech platforms had been likely to play a major purpose in the overall economy, and that lookup engines had been poised to have an outsized value. She also realized that the investigation local community was just starting to deal with issues about how to structure electronic markets and what balanced level of competition seemed like in all those marketplaces, and she was enthusiastic to aid develop that research.
But after she commenced operating at Microsoft, Athey also found out one thing she wasn’t expecting: the probable for device understanding to tackle financial problems. The creators of the Bing look for engine were conducting experiments in a way that economists only dreamed about. They were at the same time running thousands of randomized A/B tests – inquiring big quantities of “what if” queries to improved understand this kind of matters as which look for final results should rise to the major and how to run auctions for setting marketing costs on a look for website page. By comparison, she claims, economists would commonly operate one experiment in a year.
“Microsoft was utilizing an artificial intelligence program composed of hundreds of algorithms that were being all functioning together to make a research effects site,” she suggests. “That was one thing new.”
In the discipline of economics until eventually then, info mining and machine studying had been pejorative conditions for a considerably less advanced kind of data. “They had been noticed as a mechanical work out to independent cats from dogs,” she says. But at Microsoft, Athey saw an prospect to combine the computational developments from predictive device learning with statistical concept so that researchers could far better understand causal consequences not only in small business apps like the look for engine, but also in social science and economics. It was an epiphany that introduced her in a new exploration course and served outline her as one of the early tech economists.
Device Studying and Causal Outcomes
Coming out of her experience at Microsoft, Athey understood that the insights from predictive algorithms could be harnessed in new approaches by combining them with recent developments in econometrics and studies. For instance, equipment understanding algorithms could be tailored to respond to trigger-and-outcome questions in economics, this kind of as what will materialize if we change the bare minimum wage? Expand immigration coverage? Elevate price ranges? Permit two companies to merge? “Predictive device finding out can’t clear up these queries on your own, but it can assistance,” she states.
For example, Athey has utilized device finding out to seem at the influence on customers of individualized pricing, a form of price discrimination that requires charging various costs to consumers according to their willingness to fork out. Standard economics strategies would give combination options to that difficulty, she states. They would study maybe a person merchandise classification at a time, looking at need for, say, different brand names of yogurt or towels. By applying machine learning solutions to consumers’ historical order information, Athey’s exploration group can estimate personalised purchaser preferences throughout various solutions at the very same time.
Creating these predictive models of purchaser alternative in turn allows researchers to question even bigger inquiries about these types of things as what happens to charges if you use a tariff, or if generics get there on the market. “As an enter to answering these concerns, we want to comprehend how consumers make selections,” Athey suggests. And machine finding out provides that enter in a way that will allow scientists to do this operate far more efficiently, at a greater scale, and in a much more individualized way. “If you are assuming everybody’s the same, that presents diverse answers than if you assume that people today have various choices,” she states.
A Tech Economics Pioneer
Athey’s chief economist purpose at Microsoft ended in 2013, but her tenure there defined her as one of the first people today to be regarded as a “tech economist.” It is a discipline she has since assisted establish as an independent discipline by convening early conferences in the subject and mentoring various college students together that vocation path.
“Now tech economists maintain an yearly meeting that draws about 800 participants,” she states. “And we have a specialised career market place mainly because becoming a tech economist is a unique career that men and women can go after.”
Athey has also created about what it signifies to be a tech economist. “It’s partly a vocation, but it’s also a mix of unique fields of research,” she says. Tech economists research the affect of digitization on the economic system, which entails thinking about current market structure, privacy, knowledge security, fairness, levels of competition coverage, and extra, she states. “They also support produce and analyze organization designs and aggressive system, and they link the versions to knowledge to guide decisions.”
Advancing AI for Superior
At Microsoft, in addition to having an unanticipated deep dive into device mastering and AI, Athey witnessed firsthand the difficulties produced by these systems – ethical and legal problems, First Amendment problems, fairness and bias, privateness and copyright, and the prevalence of unexpected implications as men and women manipulated or gamed the program in response to sector shifts or new guidelines.
Due to the fact of these observations, Athey formulated a drive to affect the techniques that machine mastering and AI would engage in out in the entire world. When she returned to academia entire time, her first actions in that course provided assisting to prepare the start of HAI and then becoming one of HAI’s founding associate directors. “Stanford HAI was definitely developed to deal with these challenges,” she says. “We want to make AI advantageous for individuals, and we want to stay clear of all of these unintended outcomes.”
Athey also wanted to translate the profitable employs of AI from the for-revenue sector into the social affect sector. This urge led her to launch the Golub Cash Social Influence Lab at Stanford. “We’re bringing the tech toolkit to social effects programs,” she says. So, for illustration, the Social Influence Lab has executed scenario research of electronic schooling technologies to boost students’ mastering created and executed methods to focusing on educational messaging to maximize engagement of farmers created and evaluated electronic pill programs that guide nurses via counseling sufferers and designed procedures to prioritize candidates for medical trials of COVID-19 medicines.
Connecting the Dots to the DOJ
Applying equipment finding out to fascinating social challenges at the Golub Funds Social Effects Lab is a bottoms-up approach to producing transform, Athey claims. By contrast, in her new occupation as chief economist of the antitrust division at the DOJ, she will be making an attempt her hand at addressing the problems of the electronic economic climate from the top down. “Government legal guidelines and procedures impact every little thing from how level of competition performs to what mergers go by way of, to what investments people make,” Athey claims.
In transferring to the DOJ, Athey hopes to continue numerous of HAI’s initiatives to help governments adapt to an era of swiftly altering engineering, specifically all-around the use of facts in business and in authorities. “Because technology this kind of as synthetic intelligence moves so rapidly, it’s tough for the government to retain up,” she claims. “We have to determine out how all branches of federal government are heading to be prepared to manual us via a distinct age.”
It is a fantastic time for Athey to attempt her hand at governing administration get the job done, suggests Levin. “At a second when technological innovation is ascendant and marketing competitors is crucial, I simply cannot think of anyone I’d fairly have at the DOJ than Susan.”
Athey will keep on being a professor at Stanford Graduate University of Business enterprise and a senior fellow at the Stanford Institute for Economic Coverage Investigation. She will stage down from her Stanford HAI affiliate director function whilst she will keep on to be an affiliated faculty member.
Stanford HAI’s mission is to progress AI investigation, instruction, plan and follow to boost the human affliction. Find out additional.