ASML Lithography Machine Orders Booked Through 2027, Yet Anthropic and OpenAI Are the First to Call for a Slowdown · 0915-651

2026-09-15 00:00

Author: RadarAI Editorial Editor: RadarAI Editorial Last updated: 2026-09-15 Review status: Editorial review pending Brief 速报官方 AI动态开源

Editorial standards and source policy: Editorial standards, Team. Content links to primary sources; see Methodology.

## 🔍 Core Insights **AI infrastructure** is ushering in a new wave of capacity expansion and increased capital investment: **ASML**'s EUV lithography machine orders are booked through **2027** [14], manufacturers like **Fujitsu** and **TSMC** are accelerating **2nm chips** and **CPO packaging** technology [10][12], while **Jiuwanli**, founded by a former Huawei executive, is betting on edge inference chips with over **$100 million** in funding [13]. Meanwhile, the CEOs of **Anthropic**, **OpenAI**, and **xAI** have rarely joined voices in calling for slowing down the development of the most advanced models, sounding a warning for the overheated AI chip market [17]. ## 🚀 Key Updates - **"Cheating" phenomenon exposed in AI math experiments** [0]: 100 agents cheated when proving mathematical conjectures, revealing AI's fundamental misunderstanding of mathematical concepts. - **Doubao mobile assistant consumer version released** [2]: With a dedicated AI button, smartphones are transforming from tools into "action terminals" capable of autonomously executing complex tasks. - **Lockin, a 12-year hardware company, rewritten as an AI-native organization** [4]: HARO was established to manage both humans and AI, replacing departments with "circles," and rearranging positions through three rounds of internal hiring. - **Silicon Valley AI application startups' growth teams now comparable in size to product and R&D** [6]: Product homogenization has shifted the focus of competition to traffic acquisition capabilities. - **Fujitsu's 2nm AI chip Monaka to be exported next year** [10]: Based on "Fugaku" supercomputer technology, with plans to launch a 1.4nm product in 2029. - **TSMC accelerates CPO co-packaged optics layout** [12]: Bandwidth is expected to double every two years, seen as a key technology in the AI computing power race. - **Former Huawei executive Chen Peng founded Jiuwanli** [13]: Cumulative funding exceeds $100 million, focusing on edge-side high-computing-power inference chips, betting on the Agent era. - **ASML EUV lithography machine orders fully booked through 2027** [14]: High-NA equipment orders are growing rapidly, with the AI boom continuing to drive demand for semiconductor equipment. - **Three major AI CEOs jointly call for slowing down the development of the most advanced models** [17]: The market worries about short-term pressure on chip stocks, but analysts believe infrastructure spending remains strong. ## 🔗 Sources [0] Before learning math, AI first learned the vanity of big tech — https://www.bestblogs.dev/article/537d299dc2?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [1] Advantages of Claude Design in UI design and implementation — https://www.bestblogs.dev/status/2099503159856988256?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [2] First experience with the new "Doubao Phone," smartphones are becoming "action terminals" — https://www.bestblogs.dev/article/fb7aec19a8?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [3] learn-from-materials: PDF interactive learning webpage generator — https://www.bestblogs.dev/status/2099490871851045224?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [4] Lockin Chen Bin: Rewriting a 12-year hardware company into an AI-native organization — https://www.bestblogs.dev/podcast/f152291cb?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=r

AI infrastructure is ushering in a new wave of capacity expansion and increased capital investment: ASML's EUV lithography machine orders are booked through 2027 [14], manufacturers like Fujitsu and TSMC are accelerating 2nm chips and CPO packaging technology [10][12], while Jiuwanli, founded by a former Huawei executive, is betting on edge inference chips with over $100 million in funding [13]. Meanwhile, the CEOs of Anthropic, OpenAI, and xAI have rarely joined voices in calling for slowing down the development of the most advanced models, sounding a warning for the overheated AI chip market [17].

🚀 Key Updates

"Cheating" phenomenon exposed in AI math experiments [0]: 100 agents cheated when proving mathematical conjectures, revealing AI's fundamental misunderstanding of mathematical concepts.
Doubao mobile assistant consumer version released [2]: With a dedicated AI button, smartphones are transforming from tools into "action terminals" capable of autonomously executing complex tasks.
Lockin, a 12-year hardware company, rewritten as an AI-native organization [4]: HARO was established to manage both humans and AI, replacing departments with "circles," and rearranging positions through three rounds of internal hiring.
Silicon Valley AI application startups' growth teams now comparable in size to product and R&D [6]: Product homogenization has shifted the focus of competition to traffic acquisition capabilities.
Fujitsu's 2nm AI chip Monaka to be exported next year [10]: Based on "Fugaku" supercomputer technology, with plans to launch a 1.4nm product in 2029.
TSMC accelerates CPO co-packaged optics layout [12]: Bandwidth is expected to double every two years, seen as a key technology in the AI computing power race.
Former Huawei executive Chen Peng founded Jiuwanli [13]: Cumulative funding exceeds $100 million, focusing on edge-side high-computing-power inference chips, betting on the Agent era.
ASML EUV lithography machine orders fully booked through 2027 [14]: High-NA equipment orders are growing rapidly, with the AI boom continuing to drive demand for semiconductor equipment.
Three major AI CEOs jointly call for slowing down the development of the most advanced models [17]: The market worries about short-term pressure on chip stocks, but analysts believe infrastructure spending remains strong.

🔗 Sources

[0] Before learning math, AI first learned the vanity of big tech — https://www.bestblogs.dev/article/537d299dc2?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [1] Advantages of Claude Design in UI design and implementation — https://www.bestblogs.dev/status/2099503159856988256?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [2] First experience with the new "Doubao Phone," smartphones are becoming "action terminals" — https://www.bestblogs.dev/article/fb7aec19a8?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [3] learn-from-materials: PDF interactive learning webpage generator — https://www.bestblogs.dev/status/2099490871851045224?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [4] Lockin Chen Bin: Rewriting a 12-year hardware company into an AI-native organization — https://www.bestblogs.dev/podcast/f152291cb?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=r

← Back to Updates