DeepSeek is booming, will the AI Agent sector really face a downturn?

The AI Agent sector continued to accelerate its pullback today, with the exception of some meme tokens related to DeepSeek, leading framework/launchpad/agent tokens such as VIRTUAL, AI16Z, AIXBT, ARC, and other large and small tokens all recording double-digit declines.

This is closely related to the overall weakness of the current market, but a deeper observation of market discussions reveals that some investors attribute this collective decline of AI Agents primarily to the explosive rise of DeepSeek.

➤ Controversial Core

Is the iterative development of LLM going to disrupt the existing AI Agent ecosystem? Setting aside the debate over the reasoning quality of DeepSeek and other leading models, if we assume from an economically sustainable perspective that it has superior iterative potential, will the evolved DeepSeek spur the creation of new Agents, rendering existing Agents uncompetitive?

➤ To solve this problem, we need to analyze the relationship between LLM and AI Agents

DeepSeek believes that LMM (equivalent to LLM) is a core component for AI Agents in processing natural language tasks, providing language processing capabilities and helping to function in various scenarios.

Kimi points out that AI Agents' natural language understanding and generation depend on LLM, but they also integrate various abilities such as decision-making and problem-solving, and can also incorporate other tools and APIs to expand functionality. Therefore, besides LLM, other key abilities also determine the quality of AI Agents' expression and interaction.

For example, AIXBT stands out in output quality among similar Agents because it has been meticulously designed and optimized in various aspects such as prompt design, post-processing mechanisms, and context management, forming its competitive advantage.

➤ Based on the above relationship, the iterative development of LLM will not disrupt existing AI Agents

Because AI Agents can evolve through API integration with the capabilities of the next generation of LLMs, enhancing interaction quality, optimizing efficiency, and expanding application scenarios.

DeepSeek provides an API format compatible with OpenAI, creating convenience for integration. In fact, the Eliza framework developed by ai16z DAO completed support integration for DeepSeek two weeks ago.

In summary, although the market has concerns about the development of AI Agents due to the popularity of DeepSeek, from the relationship between the two and the evolution of Agents, existing AI Agents will not be easily disrupted. In the future, we should focus on how to leverage the development of LLM to drive industry progress.

$ai16z $aixbt $virtual