DeepSeek Wants AI Agents to Clock In, Boss Around, and Cut the Human Overtime 🤖

Chinese artificial intelligence company DeepSeek has revealed a design intended to make AI coding assistants more autonomous by allowing them to perform multiple tasks with less human intervention. The proposal, outlined in a paper published by DeepSeek’s DeepSeek Harness team and reported by Singapore’s South China Morning Post on Sunday, aims to reduce the need for people to coordinate and monitor multiple AI agents.

The team said managing several agents requires substantial human attention, noting: “single-task limits depend on model capabilities, while multitask limits depend on human attention.” Its paper said a person can oversee only three to five active AI agent sessions simultaneously.

DeepSeek’s proposed system would distribute tasks among agents and coordinate their work, allowing humans to focus on reviewing results and making decisions. The development comes amid intense competition among AI developers, including US companies Anthropic and OpenAI, according to the SCMP.

The report also revisited crypto trading tests involving Chinese artificial intelligence models. Cointelegraph reported last October that, according to data from blockchain analytics platform CoinGlass, Chinese AI models were outperforming US counterparts in cryptocurrency trading.

DeepSeek led one crypto trading experiment and was the only AI model to generate a positive unrealized return of 9.1%. The result surprised crypto traders, particularly because DeepSeek was developed at a fraction of the cost of its US rivals. Asia Express also highlighted a new DeepSeek ban in the US and deepening Swiss-Asia crypto ties.