By Saritha Rai, Bloomberg
DeepSeek, a Chinese AI startup that aims to disrupt Silicon Valley’s dominance, has posted multiple job openings for specialists in agentic artificial intelligence as interest in technology capable of performing tasks autonomously surges across China. The 17 available positions include roles for an Agent Deep Learning Algorithm Researcher, an Agent Data Evaluation Expert, and an Agent Infrastructure Engineer, with applicants expected to engage deeply with DeepSeek’s models in various applications such as search, content creation, and multi-modal scenarios. The company has not disclosed salary ranges.
This move signifies a shift away from traditional large language models, which have been popularized by companies like OpenAI, toward AI agents that are increasingly employed for planning, reasoning, and operating in digital environments with minimal human input. Lately, China has experienced a surge of interest in agentic technology, driven by the rapid adoption of OpenClaw, an open-source software that integrates with messaging services such as WhatsApp and WeChat.
DeepSeek, which gained global attention over a year ago with the launch of its R-1 reasoning model, is expected to reveal a new version of its flagship technology imminently. However, the Hangzhou-based startup has yet to publicly announce details regarding this new model or its release timeline. Requests for comment from DeepSeek went unanswered.
Chinese media sources suggest that the forthcoming DeepSeek V4 is optimized for coding and long-context software engineering tasks. The startup is also collaborating with domestic chipmakers, including Huawei Technologies Co. and Cambricon Technologies Corp., to enhance its models through close hardware integration.
Meanwhile, the company has been actively promoting its agentic AI capabilities, having launched DeepSeek-V3.2 in December, aiming to evolve foundational models into more sophisticated systems. The premise is that AI agents will be capable of executing various tasks for users, such as researching and booking vacations or managing bill payments and negotiations. While the tech industry has highlighted the growing capabilities of such systems, it also cautions that these technologies pose risks, as they require access to sensitive information including email accounts and banking systems.
The job listings from DeepSeek emphasize the importance of fluency in reinforcement learning techniques, which enable AI systems to improve their decision-making by learning from outcomes and adjusting based on rewards or penalties. The postings mention US coding tools like Anthropic PBC’s Claude Code and Cursor as benchmarks to surpass, indicating that DeepSeek is closely studying American systems in its pursuit to outpace them. OpenClaw is also referenced, underscoring its current relevance in China.
Companies such as OpenAI, Anthropic, and Google, along with a growing number of startups, are also pursuing similar visions of AI that extend beyond merely answering questions to autonomously executing tasks—ranging from writing and running code to navigating web interfaces and managing workflows. Last year, Meta Platforms Inc. agreed to pay $2 billion for Manus, a Singapore-based agentic AI startup with Chinese connections, reflecting the competitive landscape in this sector.
The emergence of agents capable of carrying out complex tasks with little to no human oversight could significantly change how individuals and businesses operate in various sectors. As companies like DeepSeek push the boundaries of what AI can achieve, the implications for privacy, security, and labor dynamics are likely to grow, warranting close attention from policymakers and industry stakeholders alike.
See also
Amazon’s New AI Tool Triggers 4.3% Drop in U.S. Software Stocks Amid Disruption Fears
Meta and AWS Face AI Agent Chaos, Blame Human Error for Security Breaches
Cohere Partners with Saab to Integrate AI into GlobalEye Surveillance Jets
OpenAI Shuts Down Sora Video App Following Disney Licensing Deal Amid AI Concerns
Anthropic Accuses Chinese Firms of National Security Threats Through AI Model Theft
