ChinAI #354: Industry Gossip - overdue training fee payments and overhyped embodied AI

2022-03-07 · Source: ChinAI Newsletter · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

Leiphone Detectives Vol. 4, an industry rumor series, reveals financial strains and operational challenges within China's AI sector. One leading LLM company reportedly owed a cloud vendor over 100 million RMB in training fees last year, eventually paying it back incrementally, while another LLM firm faced salary payment delays for two to three months. The report also highlights a major tech firm in Beijing overstating its embodied AI capabilities, possessing only 30 physical robots for data collection despite claiming ambitions for 1 million hours of physical data. Additionally, Chinese AI chip companies, after failed acquisition attempts by a leading *Xinchuang* system integrator, have resorted to listing on the STAR Market or Hong Kong Stock Exchange, with some exploring pivots to the software sector. The issue also includes a curated list of "ChinAI Links" covering topics like China's AI Bulletin, agentic AI controversies, the US-China trade war, self-driving car accidents, BeiDou integration, and China's AI human-like interaction service regulations.

Key takeaway

For entrepreneurs considering investment or partnership in China's AI sector, you should scrutinize financial health and verify technological claims. The prevalence of overdue payments among LLM firms and the overstatement of embodied AI capabilities suggest a need for thorough due diligence. Be wary of companies that cannot demonstrate robust physical infrastructure or stable financial operations, as these are critical indicators of long-term viability.

Key insights

China's AI sector faces significant financial and operational hurdles, including debt, overhyped capabilities, and failed acquisitions.

Principles

• Overdue payments are common in the LLM industry.
• Real-world robot data collection requires substantial physical fleets.

In practice

• Monitor LLM firms' financial stability.
• Verify embodied AI claims with physical asset counts.

Topics

• LLM Industry Finance
• Embodied AI
• AI Chip Industry
• Xinchuang Industry
• Agentic AI Regulation

Best for: Entrepreneur, Director of AI/ML, Investor, Policy Maker

Related on AIssential

• The AI Industry Is Losing — The AI industry's massive capital expenditure is unsustainable, driven by speculative demand and propping up unprofitable AI labs.
• Anthropic Customers Find Errant Charges, Auditing Startup Says — AI enterprise billing is complex and prone to errors, leading to significant overcharges for customers.
• Marc Andreessen says AI layoffs are a farce: Companies are 75% overstaffed and AI is the "silver bullet excuse" to clean house — AI is often a scapegoat for layoffs stemming from pandemic-era overhiring and organizational bloat.
• 🔮 Exponential View #572: AI’s moats, myths and moral loopholes — AI demand is surging in China despite compute constraints, while Western AI developers grapple with the paradox of job displacement.
• US Pulls List of Tech Firms Linked to China’s Military — Market sentiment is highly reactive to AI developments, impacting valuations and regulatory scrutiny across diverse sectors.
• AI Weekly Issue #493: Meta hired $145B in capex and fired 8,000 people — AI capital expenditure and workforce reductions are now explicitly linked, signaling a shift to an "AI replacement" narrative.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by ChinAI Newsletter.