(Btw, all the WWW keynotes can be found here.) Foundation models are large, pre-trained neural network trained on broad, diverse time series data with the goal of supporting many downstream tasks (forecasting, anomaly detection, causal inference, generation) across different domains. It’s like GPT-4, but for time series analysis. 🔹 Core Capabilities of a Time Series Foundation Model: Prediction — short-term, long-term, multivariate Analysis — pattern discovery, representation learning Causal Inference — learning and modeling causal relationships over time Generation — synthetic time series for simulation or augmentation Cross-domain transfer — one model works across finance, medicine, climate, etc. To use a foundation model, there are mainly two ways: Prompt-based learning vs Fine-tuning. Prompting is fast (like how you ask ChatGPT questions). Fine-tuning can be costly but would be better for narrow, high-accuracy needs. ...

On April 21, 2025, Oh! The Federal Trade Commission filed a lawsuit today against Uber, alleging the rideshare and delivery company charged consumers for its Uber One subscription service without their consent, failed to deliver promised savings, and made it difficult for users to cancel the service despite its “cancel anytime” promises. Among the many allegations, the following may not guarantee a courtroom victory, but it stands out as the most egregious issue: ...

For the paper: Price Stability and Improved Buyer Utility with Presentation Design: A Theoretical Study of the Amazon Buy Box. Ophir Friedler, Hu Fu, Anna Karlin, Ariana Tang. Accepted at The Web Conference 2025. paper pdf poster ver 1.0 Here’s the poster for WWW2025 presentation: Thanks to my advisor Hu Fu for his ideas on this.

Dianping is the monopoly rating platform in China. They have huge market. With great power comes great responsibility. Dianping is going to be in so much trouble (antitrust and regulations :)) if they don’t take active actions to harness and make good use of their power. Here’s an interesting Chinese article from the pov of restaurants: 困在评分系统里的餐饮人 驳静 (https://mp.weixin.qq.com/s/ACtVi3hphs3ml5euBcKDlQ) Some translated snippets In big cities these days, people are used to being bribed [by restaurants] with freebies in exchange for a five-star review [on Dianping]. Two decades after its founding, Dianping’s rating mechanism quietly rewritten the ground rules of the restaurant business. ...

If you’ve never used Dianping, think Yelp—but shinier, angrier, and far more dominating with the fate of restaurants in China. It’s the platform people turn to when deciding whether some ¥68 hotpot buffet would be a delicious meal or a strict no-no. Naturally, the platform holds immense power. A single low rating can haunt a restaurant’s prospects worse than bad feng shui. Restaurants, understandably, get furious when they feel wronged—especially by vague, hostile, or outright malicious reviews. In response, they often appeal to Dianping, asking for these ratings to be reconsidered and (hopefully) wiped. ...

The commercial war has officially escalated, after JD enters the takeout business. In response to JD.com’s splashy entry into their duopolized takeout delivery game, Meituan—the undisputed heavyweight of that space—is fighting back in the only language consumers care about: massive coupons. My friend pulled 23.5RMB off a 20 meal? On average, a typical order runs around ¥48, so Meituan is basically slicing half off the bill just to keep people in its ecosystem. So when JD is throwing invester’s cash at new users, Meituan bleeds too. ...

I like to think about how to spend money well, with joy, efficiency, and ideally—being from Guangzhou which is literally the most delicious region in China, I take food seriously. Lately, there’s been an interesting bit of culinary gossip about the tech giants behind our takeout orders. I’ve written before about Meituan, the heavyweight in China’s food-delivery duopoly—a platform that has perfected the dark art of rent-seeking. It squeezes the life (and margin) out of both the restaurants and the couriers. The folks ferrying my salad across the city are pulling 10 to 14-hour days for around 6000 RMB a month. Meanwhile, restaurants are bleeding out 30% of their earnings just to be listed. Efficiency is a beautiful thing—until it starts to look like feudalism in a scooter helmet. ...

The art of internet archeology led me to the distinguished lecture given by Sendhil Mullainathan, “Economics in the Age of Algorithms”, at AEA Annual Meeting that took place in Hilton San Francisco Union Square—I’ve been there, several times! The talk took place Jan. 3, 2025. The video is available https://videosolutions.mediasite.com/Mediasite/Play/cb9d64c0274d4aae98b61dd6779791b31d I’ll write about it someday :)

What Is This Poster Even For? My advisor is letting me redesign the poster (the previous version isn’t that bad… isn’t it?) for our paper “Price Stability and Improved Buyer Utility with Presentation Design” (pdf) that’s going to be presented at The Web Conference. Fine. But, to be fair, he made a solid point. Yes, there are endless design guidelines and aesthetic choices—softwares (please, don’t use Powerpoint or Word…), color palettes, font choices, the use of white space (oh the cliche ’less is more’), clever placements of shapes and arrows to guide people’s eyeballs, etc etc etc. But what about the practical question: what do we actually want the poster to do? ...

I do have some (well, a little) Theoretical-CS background. I thought I was learning math for computer science in class—I took notes, solved problems, or may have even nodded sagely once or twice. But only after when I start reading the following books did I realize… ah, this is what they were trying to teach us. Whether you’re into algorithms, theory of computation, or occasionally need to find a quick recap or intermediate result to bypass some calculations yourself, the following are the textbooks that our professors quietly (or loudly) revere—slightly biased, but just to get started. ...