Black Lake: Chinese Factories Skip Digitization and Go Straight to AI Agents

Zhou Yuxiang, head of Black Lake, a Chinese provider of cloud AI tools for factories, described in an interview with the South China Morning Post how the shift to agents is unfolding in Chinese industry. According to him, many factories do not go through the full path of traditional digitization and move straight to AI agents. The analogy will be familiar to anyone who has been to China: the country largely skipped the credit card era and went directly to mobile payments. Now, he says, the same playbook is moving beyond China's borders and onto global supply chains.

The available portion of the interview contains no figures: no customer count, no revenue, no savings metrics. Everything below should therefore be read as the vendor's claim and market logic, not as verified statistics.

The logic of the leap

Classical factory digitization is a long ladder. First come sensors and data collection, then MES and ERP, system integration, unified reference data, and data cleanup. Each step takes years and budget. Many plants get stuck halfway: the system is in place, but the data in it is incomplete, and the shop floors still run on spreadsheets and the experience of their foremen.

The agent-based approach offers a way to skip part of the ladder. An AI agent can work with what already exists: orders, correspondence with suppliers, shift reports, equipment logs. Unlike a classical system, it does not require a perfectly structured dataset, and it takes over routine work such as planning, coordination, and spotting deviations. For factories without large digital infrastructure, this path is noticeably cheaper.

Black Lake's cloud model matters here. A factory does not buy servers or hire a team of ML engineers; it plugs into ready-made tools by subscription. That is how mobile payments spread in their day: through the smartphone people already had.

What we know

• Company: Black Lake, a Chinese provider of cloud AI tools for manufacturing.
• Source of the claim: CEO Zhou Yuxiang, in an interview with SCMP.
• Thesis: Chinese factories are moving to agents, bypassing full traditional digitization.
• Analogy: skipping credit cards and going straight to mobile payments.
• Direction: exporting the model to global supply chains.
• Quantitative data in the published part of the interview: none provided.

Exporting to supply chains

The most interesting part of the claim is not the Chinese market but what the company wants to build on it abroad. Global supply chains consist of many small and mid-sized manufacturers whose digitization is also incomplete. If an agent can work on top of fragmented data, that fragmentation stops being an obstacle. Brands and large buyers could then connect suppliers to a single loop without a multi-year integration project.

The limitations are just as clear. An agent that places orders, shifts deadlines, or changes line priorities on its own makes costlier mistakes than a chatbot. A factory's data is a trade secret, and a cloud service from China raises regulatory and trust questions for a Western customer. The payments analogy does not hold fully: there, a single platform and a clear regulator existed, while manufacturing has far fewer standards.

What to check when choosing such a solution

Anyone considering agents for manufacturing should put a few direct questions to the vendor. Which actions does the agent perform on its own, and which does it only suggest to a human? How are its decisions logged, and can they be rolled back? Where is the factory's data physically stored, and who has access to it? Which metrics have been confirmed at real sites: downtime, defects, order lead times?

And most importantly: it is better to start with a narrow area, such as procurement or shift planning, where a mistake is cheap and results show within weeks. Black Lake's claim points to a real shift, but only measurable results at specific factories will confirm it.