ChinAI #158: Around the Horn (edition 5)

Greetings from a world where…

advanced Tug-of-War tactics exist

…As always, the searchable archive of all past issues is here. Please please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay support access for all AND compensation for awesome ChinAI contributors).

Around the Horn (5th edition)

A quick refresher on the structure (previous edition in ChinAI #149):

  • short preview of 10 articles related to ChinAI — published in recent weeks and sourced from scans of WeChat accounts and groups (all links go to original Chinese)

  • reply to the email and/or comment on the Substack post with the number of the article you’re most intrigued by, which will be next week’s feature translation!

  • votes by subscribers who financially support ChinAI have added weight

1) Mandatory cancellation of restrictions on external links — a good idea?

Summary: Chinese regulators recently compelled Chinese tech giants to allow users to access external URL links, ending a long-running practice of companies blocking links that direct people to pages on rival platforms. This article questions whether this regulatory action would actually encourage more competition.

Source: Deng Feng (邓峰) — Professor at Peking University and director of the school’s Institute of Law and Economics; published in Caijing Magazine (财经杂志), a leading source for business and financial news.

2) White Paper on China’s Computing Power Development Index

Summary: This white paper assesses the current level of computing power in various Chinese provinces and cities. It establishes a research framework and system of indicators for evaluating computing power development. It also develops an index for computing power development (算力发展指数).

Source: China Academy of Information and Communications Technology (中国信通院/CAICT) — a think tank under China’s Ministry of Industry and Information Technology

3) The Spring and Autumn Period of Industrial Vision

Summary: How is computer vision changing manufacturing? This article looks at global competition in this domain, with a focus on five regions: U.S., Japan, Europe, South Korea, and China.

Source: 雷锋网 (Leiphone) — a long-time source for ChinAI translations, which I sometimes liken to China’s MIT Tech Review. For more, see ChinAI #60.

4) How to improve the competitiveness of chip testing through independent ATE (automated testing equipment)

Summary: the semiconductor value chain spans many activities, including design, assembly, testing, packaging, etc. This piece digs into the equipment used to conduct such testing, and China’s prospects in this slice of the value chain.

Source: 半导体行业观察 (Semiinsights)platform that covers global trends in semiconductors. Previously recommended this article on the worsening of China’s chip talent shortage.

5) In subtitles and dubbing, AI is playing for keeps this time

Summary: Looks at how advances in deep learning are affecting the entertainment industry, including dubbing, deepfakes, and subtitle translation. Mostly focuses on developments outside of China, with some details on Chinese industry.

Source: 机器之能/jiqizhineng (Synced) — great source for ChinAI translations, often features longform articles about China’s tech industry

6) Peking University Professor Wu Xihong: bringing NLP back to the era of small data

Summary: Last month, at a forum on Language and Intelligence (which has been held in Beijing since 2016), Wu Xihong shared his personal insights on how linking physical models of vocalization with deep neural networks could serve as a new research stream for NLP. AItechtalk transcribed the speech, edited, and got Professor Wu’s approval for this published transcript.

Source: AI科技评论 (aitechtalk) — focuses on in-depth reports on developments in the AI industry and academia.

7) New Generation AI Code of Ethics

Summary: Published September 26, this code of ethics spells out 25 articles — ethical guidelines for guiding responsible AI research and development. These are developed by China’s National New Generation Artificial Intelligence Governance Expert Committee, which had previously released eight principles for AI governance and responsible AI.

Source: 科技部 (Ministry of Science and Technology)

8) How hardcore can sweeping robots get?

Summary: Looks back at the history of sweeping robots from the perspective of three eras, with a focus on the Chinese tech company Ecovacs (科沃斯).

Source: 量子位 (QbitAI) — news portal that regularly covers AI issues, similar to Leiphone

9) Report from XiaoIce Annual Conference in Beijing

Summary: The conference revealed that users have created 17 million virtual humans through the XiaoIce framework. Moreover 26% of users view virtual humans as a substitute for a real human to talk to about their concerns.

Source: 南方都市报 (NDDaily) — Southern Metropolis Daily, newspaper published in Guangzhou -- well-known for its investigative journalism.

10) 245.7 billion parameters — the world’s largest AI massive-scale model “源1.0”[Source 1.0] is released

Another week, another release of a model that claims to be China’s GPT-3. Interestingly, the article doesn’t specify who developed the model. The report does claim that the model was trained more efficiently than GPT-3.

Source: 新智元 (AI New Era) — Media site similar to Leiphone and QbitAI.

ChinAI Links (Four to Forward)

Should-read: Can Sensetime become a Chinese AI Champion

By Mercedes Ruehl, Primrose Riordan, and Eleanor Olcott, this FT article provides an insightful look at China’s leading AI startup. This snippet was particularly interesting:

“Unicorns and the other big technology companies increasingly have their AI in house,” said William Bao Bean, a Shanghai-based general partner at global venture capital firm SOSV. “Most Chinese corporates are not in the habit of paying for software, which is why a lot of clients for companies like SenseTime are government.” One former SenseTime employee told the Financial Times that in the early days it won tech customers including ByteDance, the owner of TikTok, and Hikvision, which makes CCTV cameras. But by 2019, ByteDance, was “heavily disengaging” from SenseTime in favour of building its own internal teams after working with it on AI facial filters. “It is faster as they had quicker access to the data, and they don’t need to pay SenseTime,” the person said.

Also includes comments from me.

Should-read: AI-based Emotion Recognition Tech: China’s mould for the ‘new’ socialist?

Johanna Costigan has a great piece on China’s approach to emotion recognition technology, out now in L’Atelier. She delves into an Article 19 report by researchers Shazeda Ahmed and Vidushi Marda, which covered Chinese use cases of emotion recognition, as well as pushback against and the derailment of some of the most high-profile cases. Also includes comments from me.

Should-read: Chinese AI and Gender Biases

For SCMP, Phoebe Zhang covers a report published by the Mana Data Foundation, a Shanghai-based public welfare foundation, and UN Women:

The report’s purpose was to provide concrete evidence for gender discrimination in AI algorithms so that companies learn about the problem and fix it.

The report provided gender discrimination cases in new media, search engines, open-source coding, employment algorithms and consumption models.

One critique: I don’t understand the journalistic/editorial practice of not linking to the original report in pieces like these. If journalism is the first draft of history, make it easier for people to revise and engage with your draft by linking to original sources.

Should-apply: Research Scientist, Long-term Strategy & Governance (DeepMind)

The Long-term Strategy and Governance team at DeepMind (an AI lab, subsidiary of Alphabet, primarily based in London) is aiming to hire three Research Scientists (Assistant Professor to Associate Professor level, and more senior is possible) in 2021. They are especially looking for scholars with expertise in the global governance of technology/AI. Positions will involve less independent research and publishing than typical academic jobs, but likely still some. They are considering applications now, with a hard deadline on October 15. Allan Dafoe, who co-directs GovAI, leads the team.

Thank you for reading and engaging.

These are Jeff Ding's (sometimes) weekly translations of Chinese-language musings on AI and related topics. Jeff is a PhD candidate in International Relations at the University of Oxford and a researcher at the Center for the Governance of AI at Oxford’s Future of Humanity Institute.

Check out the archive of all past issues here & please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay for a subscription will support access for all).

Any suggestions or feedback? Let me know at chinainewsletter@gmail.com or on Twitter at @jjding99