ChinAI #40: NeurIPS 18 - Taking Stock of Chinese AI Labs
|Jeffrey Ding||Dec 10, 2018|
Welcome to the ChinAI Newsletter!
These are Jeff Ding's weekly translations of writings on AI policy and strategy from Chinese thinkers. I'll also include general links to all things at the intersection of China and AI. Please share the subscription link if you think this stuff is cool. Here's an archive of all past issues. *Subscribers are welcome to share excerpts from these translations as long as my original translation is cited.
I'm a grad student at the University of Oxford where I'm the China lead for the Governance of AI Program, Future of Humanity Institute.
NeurIPS Metrics with Shanghai Jiaotong’s Acemap Tool
We looked at a lot of universities last week; let’s get back into the industry labs this week, with two translations featuring NeurIPS, which took place last week in Montreal. First one features a nifty tool by the Acemap team of Shanghai Jiaotong University, which has done some really cool scientometrics on NeurIPS papers from this year and throughout history. The big takeaways:
At this year’s conference, almost all of the top 10 institutions in terms of papers published are American, showing the absolute leading position of the United States in this field
Among Chinese institutions, Tsinghua University, Chinese Academy of Sciences, and Peking University publish the most papers in NeurIPS.
Interestingly the Acemap piece highlights (ethnic) Chinese scholars among the top 10 publishing authors, which includes some who work at American institutions, such as Eric Poe Xing, who is a professor at CMU.
Other things you can play around with in the Acemap official site: author relationship diagrams, evolution in popular NeurIPS paper topics, and useful short summaries of 1010 papers.
Roundup of Chinese AI Labs by AI Impact Factors
In Issue #31 of ChinAI we introduced Leiphone’s AI Impact Factors, a database that takes stop of Chinese AI research institutes along four lines: conference/journal papers, competitions, development projects, and corporate activities (e.g. personnel changes). Let’s check back in with the November summary of the AI Impact Factors - a few highlights:
Baidu took the top spot in the monthly AI Impact Factors for the first time, with victories by its NLP team in an AI prosthetics challenge at NeurIPS 2018, and the announcement of the Baidu Research Institute Advisory Board
Starting to notice a lot of competitions/events hosted in China and they pop up in this translation (China National Computer Congress, China Collegiate Programming Contest)
Other notable happenings: Tencent AI’s Transmart (AI-assisted translation product), Alibaba’s open source framework, “X-Deep Learning,” for big data marketing, and many more.
This Week's ChinAI Links
Chinese phrase of the Week: 一波三折 (yi1bo1 san1zhe2): full of twists and turns, used in first translation’s recounting of NeurIPS/NIPS name change
Connie Chan’s must-read piece on AI becoming the product itself rather than just the tool, features TikTok, dating app Soul, and LingoChamp. Get ready for a flood of TikTok coverage after high-profile celebrity endorsements - just remember we covered its rise 4 months ago with a translation of Li Guofei’s comparison of Bytedance (maker of TikTok) and Tencent.
Lawfare’s Sinotech section is doing some good stuff.
Jason Si (Si Xiao), head of Tencent Research Institute, on an ethical framework for AI, based on ARCC (Available, Reliable, Comprehensible, and Controllable) — Mandarin section first, English translation midway through article
Good piece by Dr. Yujia He on AI & global governance, developing resilient economies in the age of AI.
Thank you for reading and engaging.
Shout out to everyone who is commenting on the translations - idea is to build up a community of people interested in this stuff. You can contact me at firstname.lastname@example.org or on Twitter at @jjding99