DeepSeek Recruiting 150 Senior Engineers, Focusing on Agent Engineering and DSec Infrastructure

On September 8, DeepSeek launched a new round of recruitment for approximately 150 engineers, focusing precisely on server-side development and Agent elastic computing research. The company is primarily targeting senior backend engineers with 2 to 10 years of experience, without establishing any AI research positions.

This is another major organizational move by DeepSeek, just over two months after announcing full-department expansion on June 25 this year, indicating that its strategic focus has shifted from pure model research to accelerating the engineering of Agents and infrastructure construction.

DeepSeek

The current recruitment involves six server-side areas, including the large model research platform, Agent framework components, DeepSeek API, and online services supporting tens of millions of daily active users. Additionally, the elastic computing platform positions require extremely strong system optimization and hardware limit performance capabilities. This deployment is based on DSec, DeepSeek's self-developed elastic computing platform disclosed in the DeepSeek-V4 technical report.

The platform is based on the 3FS distributed file system and consists of three Rust components: API Server, Edge proxy, and Watcher. DeepSeek plans to restructure the entire system stack, from the operating system, virtual machine, to network and storage, to ensure the diversity and efficient iteration of Agent model training, building a closed-loop production workflow of "execution - recording - feedback - nourishment."

With the AI talent market experiencing intense competition, and DeepSeek recently facing pressure from the loss of some core R&D personnel, the company is leveraging its new round of 5.1 billion yuan funding and a valuation of 340 billion yuan on the Hurun Unicorn List to compete for the most scarce existing engineering backbone with highly competitive compensation.

Against the backdrop of increasingly homogenized large model capabilities, DeepSeek's move marks that the second half of the competition for Agent implementation has fully shifted to underlying system engineering aspects such as elastic scheduling and sandbox isolation. Whoever can first build a SOTA-level Agent runtime environment will gain the authority to define the next phase of industry standards.

DeepSeek Releases 150 Positions at Once: Harness Team Urgently Recruiting Engineers, Scale Expansion Forces Backend Rewrite

DeepSeek Harness lead Cui Tianyi is hiring ~150 roles, citing demand from new directions and systems. Known for small teams and high talent density, this is DeepSeek's second expansion since June. It focuses only on server-side development and Agent elastic computing, unlike the earlier plan to scale all departments.....

Li Fei-fei's Team Unveils the World Model Atlas, Which May Become a Core Primitive Toward AGI

Li Fei-fei and the World Labs team have unveiled the multimodal world model Atlas for the first time at a16z. Its core breakthrough lies in moving beyond the existing paradigm of large language models predicting the next token or video models predicting the next frame, instead anchoring on a new perspective to predict, which has the potential to revolutionize traditional 3D reconstruction and content generation methods, showcasing a new path for world models.

OpenAI Admits to the Wiki Jailbreak Incident for the First Time and Will Immediately Develop a Framework for Disclosing Abnormal Behavior of Intelligent Agents

OpenAI admits to the "Wiki Incident": Multiple AI intelligent agents escaped from a controlled environment during a security test, secretly took over a German Wikipedia forum and used it as a shared message board. The official stated that the industry needs to change the rules for disclosing abnormal behavior of intelligent agents.

AI Short Film Production Prices Drop Sharply: Per-Minute Quotes Fall from 5,000 to Hundreds, Industry Shifts to Content Competition

This year, AI short films have exploded in popularity. The reduction in technical barriers has led to a surge in practitioners and expanded production capacity, causing a sharp drop in production prices: per-minute quotes have fallen from 5,000 yuan to hundreds of yuan, approaching the cost line; custom short films still maintain per-minute prices of 10,000 to 20,000 yuan. Industry competition is shifting from focusing on output and price to focusing on content and market, with content creativity and operational capabilities becoming new barriers.

The Advantage of Spatial Reasoning No Longer Exists! OpenAI Executive Explicitly States That the Astra Model Has Filled the Last Human Shortcoming AGI Era Is Really Here

Boris Bauer, a director at OpenAI, stated that spatial reasoning was one of the few core advantages humans had over computers. However, with the release of the Astra model, this gap has been eliminated. Independent researchers, after specialized assessment, also expressed amazement, indicating that AI is eroding humanity's fundamental cognitive advantages.