China catches up

2025-06-07 · Source: Marcus on AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Corporate Strategy & Leadership, Public Policy & Governance · Depth: Intermediate, quick

Summary

China's Zhipu Z-AI has open-sourced its large language model, intensifying the "no moat" dynamic in the AI industry and potentially undermining the U.S. sector. This development, anticipated since summer 2023, challenges the viability of trillion-dollar IPOs for companies like Anthropic and OpenAI, as price wars drive token costs towards zero, making massive data center investments difficult to justify. The current AI paradigm suffers from three fundamental flaws: it is wildly inefficient, requiring extensive resources for development and operation; its unreliability makes premium pricing unsustainable; and its easy replicability fosters intense price competition and slim margins. The article suggests the broader AI race, particularly the U.S. focus on competing with China, might be misconceived, advocating instead for a shift towards developing AI for specialized fields like science and medicine.

Key takeaway

For investors and AI strategists evaluating the long-term viability of large language model companies, recognize that the "no moat" environment, exacerbated by China's open-source advancements, fundamentally challenges trillion-dollar valuations. Your focus should shift from general-purpose LLM competition to cultivating specialized AI solutions for fields like science and medicine. This approach mitigates risks from price wars and the inherent inefficiencies of current brute-force AI paradigms, fostering more sustainable innovation.

Key insights

The "no moat" dynamic, driven by inefficient and replicable AI, leads to price wars and unsustainable profits.

Principles

• Current AI models are inherently inefficient and unreliable.
• Easy replication of AI approaches fuels intense price competition.
• A zero-sum "AI race" risks global catastrophes.

In practice

• Re-evaluate investments in general-purpose LLMs.
• Prioritize AI development for science and medicine.
• Assess AI systems for long-term reliability and cost.

Topics

• Open-Source AI
• Large Language Models
• AI Industry Economics
• Geopolitical AI Competition
• AI Investment Strategy
• Zhipu Z-AI

Best for: CTO, VP of Engineering/Data, AI Architect, Director of AI/ML, Investor, Policy Maker

Related on AIssential

• Yann LeCun warns AI labs like OpenAI and Anthropic face a "big bubble explosion" — AI labs face financial unsustainability due to high operating costs and investor subsidies, risking a market bubble.
• The Next US-China AI Battle Is Over Compute — and China Is Spending Big — The US-China AI rivalry is intensifying, with compute infrastructure becoming the primary battleground due to China's significant investment.
• My bets on open models, mid-2026 — The open vs. closed AI model race is primarily an economic contest, not solely a technical capability one.
• ChinAI #359: DeepSeek's "Huawei-like" Mission in AI — DeepSeek's soaring valuation reflects its strategic role in China's drive for AI self-sufficiency amidst external tech restrictions.
• The AI Revolution Will Not Be Monopolized: How open-source beats economies of scale, even for LLMs — Open-source development is posited as a critical force to prevent monopolization of the AI revolution, particularly for LLMs.
• Why things will eventually fall apart — The AI market, especially for LLMs, lacks moats, leading to commoditization, price wars, and unsustainable valuations.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Marcus on AI.