State of the Claw
summary
Five months in, OpenClaw claims to be the fastest growing open source project in history with ~30,000 commits, ~2,000 contributors and thousands of PRs, and it faces real operational strain: 1,142 security advisories (99 labeled critical, ~60% closed), repeated sandbox escapes found via Nvidia’s NemoClaw sandbox escape + Codex security testing, and supply‑chain hits through Slack/MS Teams. A foundation forms to hire maintainers, integrations with Nvidia/Microsoft/Red Hat/Tencent ship, and work goes into dreaming and modular plugins to make the system usable and safer.
AI-written summary · about
chapters (16)
- 0:00Project Growth and Statistics
- 2:23Management Challenges and the OpenClaw Foundation
- 3:47Addressing Security Advisories and Vulnerabilities
- 10:33Misinformation and Media Fearmongering
- 14:50The Burden of Open Source Maintenance
- 16:12OpenAI Involvement and Future Independence
- 18:57Audience Q&A Begins
- 19:53OpenClaw's Relationship with OpenAI
- 22:28The Importance of Open and Local Models
- 24:57Coding Workflow and Agent Interactions
- 28:28Defining 'Taste' in AI Development
- 30:31Developing Personality for AI Agents
- 33:22Future Vision: Ubiquitous Agents and Smart Homes
- 35:58Addressing Prompt Injection Risks
- 38:33Future Vision: Implementing 'Dreaming' and Modularity
- 40:24Life as a Maintainer and Future Skills
related talks
31m📙 From Assistant to Assassin: Weaponizing An OpenClaw Vulnerability to Achieve 1-Click RCE 🖥️
25mOpen Source Projects and AI Security
12mCNCF Projects Update
40mOpen Source Is Not the Same Anymore
22mKeynote: Opening Remarks
16mPerformance improvements: exposing hidden security fixes in Open Source
44mSecuring Open Source Software - End-to-End, at Massive Scale, Together
36mState of Open Source in 2024