Discover more from ChinAI Newsletter
ChinAI #174: Five Factions Competing for China's Industrial Internet
Plus, another example of "techlore"-style writing
Greetings from a world where…
…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).
Feature Translation: All the Land of the Industrial Internet, Five Rising Chinese Forces
Context: Previously, I’ve identified China’s efforts to build an “Industrial Internet” as an important, relatively neglected topic (ChinAI #70). The Industrial Internet of Things = a system of industrial devices integrated with information and communications technologies so as to facilitate advanced analytics, machine-to-machine coordination, and other functions. In the words of this week’s Leiphone article (original Chinese): “The Industrial Internet is actually everywhere – as big as China’s international strategic position, as small as you and me. It stands there prominently among the seven main items of the “new infrastructure” concept, and it has become a celebrity in China’s macro and micro strategies for four consecutive years.”
Key Takeaways: The article groups the companies involved in building China’s industrial Internet into 5 factions (see table below):
Possessing homefield advantage with their knowledge of production and manufacturing processes, traditional manufacturing companies have led the way in building industrial Internet platforms (e.g., CASICloud’s Indics, Haier’s COSMOPlat, and and Foxconn’s FII Cloud). Industrial companies account for 46% of Chinese industrial Internet platforms, per CAICT data from 2018.
Startups slicing up specific production scenarios to apply AI. From the article: “When many AI companies devote themselves to the security field, where it is easier to to produce results, and autonomous driving, where it is more gimmicky, these startups initially chose to set up camps on this more difficult land. As a start-up, the courage to bet on industrial applications is deserving of people’s admiration.”
Internet giants have stepped up their efforts in recent years. An intriguing narrative has developed: Over the past decade, internet giants have been the biggest early adopters and beneficiaries of the informatization revolution, but now the battle for traffic has peaked. The industrial Internet gold mine is yet to be unearthed. Who will “seize the second half?” Full translation contains good details about what the BAT companies working on in the industrial Internet field.
Before describing the last two factions (industrial software companies, ICT hardware companies), I want to note that this piece is a great example of what I call “techlore” (ChinAI #98): epic poem-like articles by development bloggers in which Chinese tech companies wage battle over the commanding heights of the economy.
When talking about one Chinese company’s efforts in electronic design automation: “For Huada Empryean, the three international giants (Cadence, Siemens EDA, and Synopsys) occupy the absolute dominant position in the EDA market, and Huada Empyrean is clinging to its position, trying to open a hole in the EDA Iron Curtain.” Ten years of technical accumulation is just the starting point for the industrial Internet, as one saying goes. This is even more true in the field of industrial software.
In the ICT manufacturers camp, Huawei and Inspur have used their head start with government cloud applications to also implement their industrial Internet solutions.
Overall, grasping the distinctions between these five types of companies is useful for understanding dynamics in areas beyond the industrial Internet.
FULL TRANSLATION: All the Land of the Industrial Internet, Five Rising Chinese Forces
ChinAI Links (Four to Forward)
Must-read: Choices, Risks, and Reward Reports
Co-authored by Thomas Gilbert, Sarah Dean, Tom Zick, and Nathan Lambert, this UC Berkeley Center for Long-Term Cybersecurity present a case for “reward reports” that help us actively specify and evaluate reinforcement learning systems. They write: “Our broader vision is that, in safety-critical domains, the role of an algorithm designer matures to be closer to that of a civil engineer: a technical expert whose credentials and demonstrated skills are trusted to oversee critical social infrastructure, and are worthy of certification by public authorities.”
Kayla Goode and Dahlia Peterson, in an op-ed to The Hill, outline a wise approach to AI education policy based around more coordination (e.g., cross-state collaboration and shared assessment metrics for computer science education) and investment (e.g., by encouraging more educators to gain skills in teaching computer science). The op-ed draws on their excellent research in this area.
By Jennifer Conrad and Will Knight, for Wired, provide an informative overview of China’s recent regulations on AI, with good comparative analysis on what’s happening in other countries.
Natasha Gilbert and Max Kozlov, for Nature, cover the end of the China Initiative, with appropriate attention to the long-lasting harm this policy has done. We talk a lot about the good that different types of policy proposals could do. We don’t talk enough about not doing harm. Thought exercise: What the U.S. could have gained from not implementing and continuing something like the China Initiative might outweigh all the benefits from any tech policy that the Biden administration implements. This is why “don’t do stupid shit” is such an important basic approach to policymaking.
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 postdoctoral fellow at Stanford's Center for International Security and Cooperation, sponsored by Stanford's Institute for Human-Centered Artificial Intelligence.
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 firstname.lastname@example.org or on Twitter at @jjding99