UPDATE: Anthropic researchers are calling for a major shift in the AI industry, urging companies to stop building numerous AI agents and instead focus on developing essential “skills.” This groundbreaking approach promises to enhance the effectiveness of AI in real-world applications.
During the AI Engineering Code Summit last month, Barry Zhang and Mahesh Murag emphasized that the future of AI lies not in more agents, but in equipping a general agent with specialized skills. “The agent underneath is actually more universal than we thought,” Zhang stated in a clip released Monday.
The tech industry has rapidly pursued AI agents for various tasks, but researchers argue that these agents often “lack expertise” and fail to grasp important context. By implementing skills—organized collections of procedural knowledge—AI can perform tasks with greater consistency and efficiency.
In just five weeks since launching, users, including those from Fortune 100 companies, have created thousands of these skills, which are being treated as internal playbooks for AI. Murag highlighted that companies are using these skills to teach agents about their organizational best practices, revolutionizing how AI integrates into business environments.
“Skills help fill in those gaps by giving agents domain knowledge and reusable workflows,” Zhang explained, underscoring the immediate need for this innovative approach.
The rise of AI agents has been touted as a game-changer for office work. OpenAI CEO Sam Altman remarked in June that AI agents are already taking on tasks typically handled by junior employees, transforming workplace dynamics.
However, some industry experts caution that the excitement surrounding AI agents may be overblown. Guido Appenzeller from a16z noted that many startups are merely rebranding existing technologies as agents to justify higher prices, raising questions about the true value of these solutions.
As the conversation around AI continues to evolve, the focus is now on how skills can redefine the role of AI agents in various sectors. Companies and developers are urged to pivot towards this new paradigm, making it critical for stakeholders to stay informed on these developments.
Next steps: Watch for further updates as more organizations adopt skills-based frameworks for AI, potentially reshaping the future of work and productivity.
