红杉资本合伙人放话：从会聊到会干，2026年AGI已经来了

Core Insights - Sequoia Capital asserts that Artificial General Intelligence (AGI) has already arrived, emphasizing its presence in "long-cycle intelligent agents" that can operate autonomously for extended periods [1][13] - The definition of AGI is simplified to the ability to "understand and solve problems," focusing on practical applications rather than philosophical debates [5][4] Group 1: Definition and Understanding of AGI - The definition of AGI has been debated for years, but Sequoia partners propose a straightforward definition: AGI is the ability to "understand and solve problems" [5][4] - An effective intelligent agent requires knowledge, reasoning capabilities, and the ability to iterate and learn from mistakes, which are now being integrated into AI systems [5][6] Group 2: Practical Applications of AI - A specific example illustrates AI's capabilities: an intelligent agent successfully identified a suitable candidate for a job in just 31 minutes, showcasing its ability to navigate complex tasks autonomously [3][6] - The AI agent performed a series of sophisticated operations, including analyzing resumes, reviewing industry presentations, and assessing social media activity to find the best candidate [6][7] Group 3: Technological Advancements - The emergence of "long-cycle intelligent agents" allows AI to maintain focus and manage tasks over extended periods, significantly enhancing its operational capabilities [9][10] - The ability of AI to complete long-duration tasks is improving exponentially, with capabilities doubling approximately every seven months, indicating rapid advancements in AI technology [9][10] Group 4: Future of Work with AI - The concept of "employing" AI agents is becoming feasible, with specialized AI models emerging for various professional roles, such as AI doctors, lawyers, and researchers [11][12] - The transition from AI as a conversational tool to an active executor of tasks is expected to reshape work dynamics, with humans shifting from direct execution to managing AI teams [11][12]