I always like attending job talks. You learn something, be awed by cutting edge research, and from time to time, reflect on skills of academic presentations… Today’s topic is Multi-Objective Optimization (MOO). When sometimes for an optimization problem there are two or more conflicting objectives (e.g. fairness vs. efficiency), it is not guaranteed that a single solution simultaneously optimizes each objective. Naturally, we consider Pareto Efficiency as a major concept for solution—a solution where none of the objective functions can be improved in value without degrading some of the other objective values. By the way, quick overview for Pareto: ...

When will we get a “XXX is all you need” paper published in an economic journal? (This comes from an ML paper ‘Attention is All You Need’, which introduced ‘Transformers’—the tech behind ChatGPT. Since then, there’s been a flood of papers claiming ‘X is all you need’ especially in computer science papers… ) source: Types of Machine Learning Papers Reddit

PhD in different fields is vastly different. A very successful computer science PhD students is often expected to graduate with at least 10 papers, while economic PhDs need only one. Math PhDs can even do without a paper. For anyone who is applying for or considering a PhD in economics, this book is a good start: Getting a PhD in Economics by Stuart J. HillmonConsidering a graduate degree in economics? Good choice: the twenty-first-century financial crisis and recession have underscored the relevance of experts who know how the economy works, should work, and could work. However, Ph.D. programs in economics are extremely competitive, with a high rate of attrition and a median time of seven years to completion. Also, economic professions come in many shapes and sizes, and while a doctoral degree is crucial training for some, it is less beneficial for others. How do you know whether a Ph.D. in economics is for you? How do you choose the right program—and how do you get the right program to choose you? And once you’ve survived years of rigorous and specialized training, how do you turn your degree into a lifelong career and meaningful vocation? ...

Economic is a very broad field. But this might be the most chic way of being an economist in 2025: a designer. Presidential Address: The Economist as Designer in the Innovation Process for Socially Impactful Digital Products Susan Athey | American Economic Review, April 2025 This paper provides an economic perspective on data-driven innovation in digital products, focusing on the role of complex experiments in measuring and improving social impact. The discussion highlights how tools and insights from economics contribute to each stage of the innovation process. Key contributions include identifying problems, developing theoretical frameworks, translating goals into measurable outcomes, analyzing historical data, and estimating counterfactual outcomes. The paper also surveys recently developed tools designed to address challenges in designing and analyzing data from complex experiments. ...

Today is the Qingming Festival (Tomb-Sweeping Day)—it’s a celebration of spring and memory. During Qingming, families visit the tombs of their ancestors to clean the gravesites and make ritual offerings to their ancestors. Or just go out for a spring hiking in nature. At the Orchid Pavilion at Shaoxing City I’m not a big fan of classical Chinese literature—I know it’s deep and awesome but please that’s just not my thing. But gladly if you went through high school in China and passed the college school entrance examination with the 10 million students every year, you are able to recite well the Orchid Pavilion Preface (and a few other very famous classical Chinese literature pieces). ...

18 Ariana Grande studio acappellas—it’s like listening to the angel sing for 90 minutes straight!!! Buy it. But it. Link to official website for the album. Ariana Grande recently released an official a cappella version (e.g. vocals only, no instrumentals) of her whole eternal sunshine deluxe album. Ariana is known for her vocal stacking—on her previous album, Sweetener, the track “get well soon” famously maxed out Pro Tools’ limit of 256 vocal layers. Her producer was like, “Sorry, we can’t add another harmony—you broke the software.” Her new a cappella album is available digitally on her website for just $9.99, which honestly is a pretty reasonable price for such a meticulously crafted vocal showcase. Highly recommended for music geeks. ...

When an AI server says “I’m here to help!"—they aren’t here to help you—it’s there to tire you out until you accept your fate or log off in frustration. Companies today, in their never-ending quest to cut costs, have increasingly turned to AI-powered solutions for customer service—so they (i) hire fewer human agents, and (ii) make it frustratingly difficult for you to reach an actual person. Back in the old days, AI customer service was nothing more than keyword-triggered auto-responses and simplistic, rule-based bots. Today, however, these systems—powered by sophisticated large language models—have become annoyingly competent at handling basic inquiries. So now, companies aren’t just deploying AI to streamline services; they’re intentionally adding friction to prevent customers from easily reaching human agents. ...

Something fun for April 1st: Goofy (and economically spicy) IP bleaches—aka, copycats! Miu Miu vs. Zara Miu Miu is high fashion. Zara is high speed. One sells couture dreams; the other sells dreams on clearance. For trendy (and budget-aware) fashion girls, Zara is a cheat code—you can literally recreate an entire Miu Miu look for about 5% of the price. TikTok is full of tutorials showing how to turn YSL, Miu Miu, or Prada show into a Zara haul. ...

Today at Columbia Business School’s DRO PhD open house, a friend of mine from SJTU is really into something called “experimentation”—and introduced it to me. I found is somewhat echos online learning. But not entirely the same. Experimentation in OR/OM means actively exploring different actions to learn about unknown factors, rather than passively using assumed parameters to optimize a system. This concept is central to online learning problems like multi-armed bandits and reinforcement learning. Algorithms balance exploration vs. exploitation to maximize cumulative reward. ...

Last summer, my then-boyfriend—not an economist but, as it turned out, a very rational kind of asshole—drove me to the airport to San Francisco. As I trying to connect up Apple Music in the car, I frustratingly struggled with his ancient iPhone 8. He muttered that “the new one is already ordered.” I nodded. Fine. Weeks later, I landed back from Yale, and he was there to pick me up again—with the same cracked iPhone 8. He explained that he had read online that the iPhone 16 would be released soon, with significantly better performance and the same launch price as the current iPhone 15. So, being the rational agent he is, he canceled the order and decided to wait. ...