Stanford drops 45-min SpaceX lecture on how Grok Bot's always-on agent harness works
Hesamation · x · 2026-10-16
Stanford CS146S published a 45-minute lecture from SpaceX on how Grok Bot works, framed as a harness-engineering masterclass. Topics: what changed in the models to enable always-on agents, the architecture of an always-on agent, harness design, and context engineering for long tasks.
A slide explains why the agent runs server-side: thin client / thick server, work continues when the laptop closes, seamless phone-desktop handoff, shipping code without App Store review, offloading heavy work, a fully decoupled agent loop, easier scaling, and the point that agents running for months can't live on-device.
More from coding & agent
- Write the Agent Loop by Hand First: What LangGraph Actually Adds — m3m3o · 2026-10-16
- $20-plan users have no idea how far AI agents have come in 10 months — realmeetjames · 2026-10-16
- Google open-sources AX, a Kubernetes-style orchestrator for billions of AI agent tasks — thisdudelikesAI · 2026-10-16
- Tencent Hunyuan open-sources Hy Memory 2.0: model, SDK and MemHub for agent memory — TencentHunyuan · 2026-10-16
- ERP•AI launches headless SaaS with $0 seats that bills machines: $1 buys 4,000 record writes — rohanpaul_ai · 2026-10-16
- FDE-Gym benchmark shows huge gap on real AI delivery tasks: top model scores 0.563 — DeepLethe · 2026-10-16