Podcast Summary: 十字路口Crossing
Episode: OpenClaw 之后，谁将定义主动式 AI 的新战场？｜对谈 AirJelly 黄柏特
Host: Koji
Guest: 黄柏特 (AirJelly 联合创始人兼 CEO)
Date: March 25, 2026

Overview

In this episode, Koji sits down with AirJelly CEO 黄柏特 for an in-depth discussion on the emergence of “主动式 AI”（Proactive AI）following the release of OpenClaw. The conversation centers on the next competitive frontiers of AI, how companies and individuals can anticipate (rather than merely react to) new capabilities, and who is likely to define the rules of this rapidly shifting battlefield. The episode is rich in both technical perspectives and strategic insights, with a candid look at the evolving relationship between AI, product design, and user empowerment.

Key Discussion Points and Insights

1. What is 主动式 AI? (03:15-12:42)

• 黄柏特 distinguishes “主动式 AI” from traditional, reactive AI. Rather than only responding to user prompts, proactive AI fundamentally anticipates, initiates, and shapes user interactions.
• 黄柏特 (04:17):
  “以往我们说的 AI，更多是一种工具，用完即止。主动式 AI，它会主动提出建议、协调资源，甚至在你还没想到的时候介入。”
  (“Traditional AI functioned more like a tool, acting only when used. Proactive AI will give suggestions, coordinate resources, even intervene before you consciously realize the need.”)
• They discuss AirJelly’s early experiments with “主动调度”（Proactive Coordination) and how it’s changing user expectations.
• Koji asks about user trust and boundaries: how can an AI be proactive without being intrusive?

2. OpenClaw的冲击：谁能定义标准？ (12:43-24:30)

• OpenClaw, as the latest open-source proactive AI framework, democratized access to cutting-edge tools, sparking a new wave of developers and startups.
• 黄柏特 (15:02):
  “跟以前GPU、深度学习爆发的阶段很像，现在是应用和框架百花齐放。”
  (“Much like the explosion of deep learning post-GPU, we’re seeing a flourishing ecosystem of applications and frameworks.”)
• Challenges discussed: fragmentation, lack of standards, and the risk of early lock-in by big platforms.
• Koji and 黄柏特 analyze potential “gatekeepers”—will it be open-source communities, major cloud platforms, or unexpected new players?
• The question of who will “define the new rules of the game” is left intentionally open, as both guests agree it is still unfolding.

3. 产品哲学：“定义主动，而非被定义” (24:31-35:20)

• Focus shifts to product positioning and philosophy. 黄柏特 advocates for empowering users to define their own boundaries with proactive AI, rather than letting the tool dictate behavior.
• 黄柏特 (28:47):
  “如果主动型AI帮你变得更好，而不是让你被动接受，这时才有意义，而且这种主权意识非常重要。”
  (“Proactive AI should help you become better, not turn you into a passive recipient. This sense of agency is crucial.”)
• Challenges: balancing personalization with privacy, ensuring transparency in AI decision-making.

4. 未来展望 — 从工具到“AI伙伴” (35:21-48:00)

• Discussion on how proactive AI will blur the lines between tool and companion.
• Koji (37:59):
  “你觉得未来人们会不会对AI有情感投入？比如说把AI当朋友？”
  (“Do you think people in the future will develop emotional attachments to their AI, perhaps see them as friends?”)
• 黄柏特 predicts that users will increasingly blend emotional and practical relationships with AI, especially as long-term interactions generate unique shared experiences.
• Emphasis on “co-evolution” — as users grow, so must their AI companions.

5. Advice to Founders and Builders in the AI Era (48:01-53:30)

• 黄柏特 encourages entrepreneurs to think beyond current paradigms:
  • Don’t imitate; instead, “define your own questions and problems, then let AI help you answer them.”
  • Collaboration is key: open ecosystems will likely yield the most sustainable innovation.
• 黄柏特 (51:10): “大胆做梦，小心落地。”
  (“Dream boldly, execute carefully.”)

Notable Quotes & Memorable Moments

• On proactive AI’s potential:
  “它像是一种新的基建，让所有人都能获得‘主动型能力’。”
  (“It’s like new infrastructure, granting everyone proactive capabilities.”) — 黄柏特 (10:52)
• On the uncertainty of AI’s next ‘gatekeeper’:
  “说到底，主导权永远在变化。看谁能定义交互方式、谁能赢得用户信任。” — Koji (23:44)
• On designing with agency:
  “主权感是所有创新的起点。” — 黄柏特 (32:14)

Timestamps for Important Segments

• 03:15 – What is proactive AI & its distinguishing capabilities
• 12:43 – OpenClaw’s impact and ecosystem shakeup
• 24:31 – Product design and user agency with proactive AI
• 35:21 – The future: will AI become a true ‘companion’?
• 48:01 – Advice for founders in the age of proactive AI

Overall Tone and Takeaways

The dialogue is forward-looking, candid, and thoughtful—characterized by optimism for AI’s transformative power, but also a call for caution and responsibility. Listeners walk away with a nuanced understanding of the proactive AI landscape, its unresolved questions, and a resonant message about the importance of user agency and ethical design.