AI in Combat: When Machines Decide Faster Than Humans
What happens when AI systems process battlefield data faster than humans can evaluate it? Exploring decision compression, autonomous targeting, and the future of military AI governance.
KubeCon Europe 2026 is in Amsterdam this March, and I am bringing something unusual in my backpack: a Raspberry Pi running OpenClaw, my personal AI agent that manages my website, monitors my infrastructure, and handles routine DevOps tasks autonomously.
While everyone at the conference will be talking about AI agents in theory, I will be showing one that actually runs β on a $60 single-board computer.
OpenClaw is an open-source AI agent framework that turns language models into persistent, tool-using assistants. Unlike chatbots that forget everything between sessions, OpenClaw maintains memory, runs on a schedule, and integrates with real infrastructure.
Think of it as the difference between asking ChatGPT a question and having a junior DevOps engineer who never sleeps.
The cloud-native community talks a lot about scaling up. I am more interested in scaling down.
Running OpenClaw on a Pi proves several points:
# Install OpenClaw on a Raspberry Pi
curl -fsSL https://get.openclaw.ai | bash
# Configure with your preferred model
openclaw config set model claude-sonnet
# Start the gateway
openclaw gateway startThe entire setup takes about 10 minutes.
At KubeCon, I will demo the real tasks my OpenClaw agent handles daily:
My agent manages lucaberton.com β this very website. It writes blog posts, fixes SEO issues, updates event listings, and commits changes to GitLab. The KubeCon Europe 2026 Side Events Guide you might have seen? My agent helps maintain it.
# HEARTBEAT.md β agent checks these periodically
- Check GitLab CI pipeline status
- Monitor website uptime
- Review Cloudflare analytics
- Check for new Ahrefs alertsThe agent generates technical content, optimizes metadata, fixes broken links, and manages cross-site SEO across my network of educational websites including Ansible Pilot, Terraform Pilot, and CopyPasteLearn.
βββββββββββββββββββββββββββββββ
β Raspberry Pi 5 (8GB) β
β βββ OpenClaw Gateway β
β βββ Memory Store (files) β
β βββ Git (workspace) β
β βββ Node.js runtime β
ββββββββββββ¬βββββββββββββββββββ
β HTTPS
βΌ
ββββββββββββββββββββββββ
β LLM API (Claude) β
β Discord / Telegram β
β GitLab API β
β Web Search β
ββββββββββββββββββββββββThe Pi handles orchestration, memory, and tool execution. The heavy reasoning happens in the cloud via API calls. This hybrid approach gives you the best of both worlds: local control with cloud-scale intelligence.
If you are attending KubeCon Europe 2026 in Amsterdam, I would love to chat about:
You can find me at Cloud Native Rejekts on March 21-22 where I am serving as Master of Ceremonies, and throughout KubeCon week at various side events.
Want to try it yourself? OpenClaw is open source and runs on any Linux machine:
# Quick start
curl -fsSL https://get.openclaw.ai | bash
openclaw gateway start
# Connect to Discord
openclaw config set discord.token YOUR_BOT_TOKEN
openclaw gateway restartCheck the OpenClaw documentation and join the community on Discord.
KubeCon Europe runs March 24-26 at the RAI Amsterdam, with co-located events on March 23. Check out my complete side events guide to plan your week.
If you spot someone with a Raspberry Pi at a conference party, that is probably me. Come say hi β my agent might even respond faster than I do.
AI & Cloud Advisor with 18+ years experience. Author of 8 technical books, creator of Ansible Pilot, and instructor at CopyPasteLearn Academy. Speaker at KubeCon EU & Red Hat Summit 2026.
What happens when AI systems process battlefield data faster than humans can evaluate it? Exploring decision compression, autonomous targeting, and the future of military AI governance.
AI is accelerating military decision-making from hours to seconds. What happens to human judgment when machines process targets faster than commanders can think?
From decision dominance to autonomous targeting, AI is transforming military operations. Are humans still meaningfully in control of life-and-death decisions?