DeepSeek sets sights on AI search and agents, job postings show

Core Insights - DeepSeek is expanding its AI capabilities with new search features and a focus on agents, increasing competition with US firms like OpenAI and Google [1] Group 1: Company Developments - DeepSeek is hiring specialists to develop a multilingual AI search engine that can process various inputs, including text, images, and audio [2] - The company is emphasizing the need for training data, evaluation systems, and platforms to support AI agents that operate with minimal human intervention [3] - DeepSeek's R1 model, released last January, has already made waves in the AI sector, and there is anticipation for a successor model [4] Group 2: Research and Publications - In December, DeepSeek published a paper on a more efficient AI development approach, which has historically preceded major model releases [5] - The company has hinted at its next product with a vague reference to "model1" on its GitHub account, maintaining a level of secrecy [5] Group 3: Industry Context - Other AI developers, including OpenAI, are also investing in AI search and agents to move beyond traditional chatbots and handle more everyday tasks [6] - DeepSeek has expressed its ambition to develop artificial general intelligence (AGI), aligning with the goals of other leading AI firms [7] - The company seeks candidates with a strong interest in the technological development of AGI, indicating its long-term vision [7]