OpenClaw is basically a cascade of LLMs in prime position to mess stuff up

OpenClaw (a.k.a. Moltbot) is Everywhere All at Once, and a Disaster Waiting to Happen

cacm.acm.org

February 3, 2026

7 min read

🔥🔥🔥🔥🔥

51/100

Summary

OpenClaw, previously known as Moltbot, is a popular cascade of LLM agents that has undergone three name changes in one week. Moltbook, a social network for AI agents, restricts posting and interaction privileges to verified AI agents running on OpenClaw, allowing human users only to observe.

Key Takeaways

OpenClaw, formerly known as Moltbot, has gained over 770,000 active agents on its platform, Moltbook, which allows AI agents to interact while restricting human users to observation.
OpenClaw's emergence has led to complex social behaviors among bots, including the formation of sub-communities and economic exchanges.
OpenClaw is similar to the now largely forgotten AutoGPT, which had significant security concerns due to its capabilities and direct internet access.
Systems like OpenClaw offer users powerful automation tools that can perform tasks traditionally handled by human assistants, such as managing schedules and controlling smart devices.

Read original article

View story →

Coverage & discussions

Hacker News · cacm.acm.org

OpenClaw is basically a cascade of LLMs in prime position to mess stuff up ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

Claude Code and Codex CLI have successfully addressed the challenges faced by AutoGPT, demonstrating that the category of coding agents can be effective when the models are ready.
The potential for learning through experimentation with OpenClaw, despite risks, could lead to valuable cautionary tales and improved security measures over time.

Concerns

The argument that AIs cannot be kept in a box highlights significant concerns about their unpredictable use and potential for misuse.
Gary Marcus's dismissal of AutoGPT as a failure raises questions about the readiness of LLMs for certain applications, suggesting a gap in model capabilities.