Dylan Patel 称"持久智能体"是 AI 应用的下一个突破口
SemiAnalysis 的 Dylan Patel 说他们 45% 的 AI 开销已经花在持久智能体上，还点评了 Claude Tags 和 OpenAI 的 Slack agent 都不行，主力在用 Perplexity Computer。想看一线用户怎么真金白银投票的，这条值得刷一下。
SemiAnalysis 的 Dylan Patel 提出"Persistent Agents"（持久智能体）这一概念，认为它是 AI 用量的下一个增长点。他表示自己团队在 AI 上的支出中已有 45% 花在持久智能体上。在具体产品选择上，他认为 Anthropic 的 Claude Tags 体验不佳，OpenAI 的 Slack agent 表现更差，团队主要使用 Perplexity Computer，该产品调用多种模型，其 harness（调度框架）表现出色。
这条观点来自 LangChain 创始人 Harrison Chase 的转发认可，他认为"Persistent Agents"是个很好的命名。持久智能体与一次性对话机器人的区别在于能长期驻留、持续执行任务，这类用法正在替代传统订阅式聊天场景，成为企业 AI 支出的新去向。
Persistent agents is a great noun Dylan Patel @dylan522p Persistent Agents are the next unlock of AI usage Our spend on AI is now 45% persistent agents Unfortunately for Anthropic, Claude Tags sucks OpenAI's Slack agent is even worse We primarily use Perplexity Computer, which uses a lot of different models Perplexity's Harness mogs 🔗 View Quoted Tweet📊 12 ⚡
This report is categorized under AITOP's "Products" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Persistent Agents, Perplexity, Anthropic, OpenAI.
Source attribution: curated from Harrison Chase (@hwchase17). Cross-source corroboration: 11 independent sources have covered this development (including Aravind Srinivas, Dylan Patel (SemiAnalysis), 宝玉), confirming industry-wide signal strength.
