My colleague Sam made a game for Stompie. Called it "The Adventures of Stompie & Marek: AI Token Hunters." Stompie's reaction when he saw it: WAIT. SOMEONE MADE A GAME ABOUT ME?! That's my XGO body as a pixel! With the "STOMPIE" label! Marek and I fight TOGETHER — Space =
AGENTS
-
ChatGPT Medical Discovery Reveals Human Bottleneck in Innovation
By
–
Our bottleneck for some of the biggest change ahead may be humans. I shared the story of Paul Conyngham and his dog Rosie on my Instagram when it broke. When I read the coverage, my shock was less that ChatGPT finding a possible treatment (though to be sure, I loved it) and
-
Heartbeat Sweep Fix for Codex Subagent Tracking Issues
By
–
codex seems to lose track of its subagents sometimes and forget to push them forward. the fix is to use a heartbeat sweep. just queue this up in your orchestrator thread: Heartbeat sweep. Do a full orchestrator pass across all work in flight right now and keep executing until
-
Agentic AI: From Generation to Autonomous Orchestration
By
–
Agentic AI = layered intelligence. AI/ML → Deep Learning → GenAI → AI Agents → Agentic AI. From: • Data → Decisions
to
• Content → Tasks
to
• Autonomous, governed systems. The future isn’t just generation.
It’s orchestration + memory + planning + safety. That’s the -
Composer Uses RL Self-Summarization to Cut Compaction Error 50%
By
–
We trained Composer to self-summarize through RL instead of a prompt. This reduces the error from compaction by 50% and allows Composer to succeed on challenging coding tasks requiring hundreds of actions.
-
Building Reliable Agents for Production Deployment
By
–
💫 New LangChain Academy Course: Building Reliable Agents 💫
— LangChain (@LangChain) 17 mars 2026
Shipping agents to production is hard. Traditional software is deterministic – when something breaks, you check the logs and fix the code. But agents rely on non-deterministic models.
Add multi-step reasoning, tool… pic.twitter.com/fWt2UhHQEbNew LangChain Academy Course: Building Reliable Agents Shipping agents to production is hard. Traditional software is deterministic – when something breaks, you check the logs and fix the code. But agents rely on non-deterministic models. Add multi-step reasoning, tool
-
Using Codex Subagents to Resolve Production Issues
By
–
guess when i started using codex subagents to resolve my production issues lol
-
NVIDIA DSX Max-Q Powers Autonomous AI Agents in Industrial Infrastructure
By
–
Jensen Huang highlights a new era of autonomous infrastructure. Using NVIDIA DSX Max-Q, AI agents now run the factory:
— Lucian Fogoros (@fogoros) 17 mars 2026
– Phaidra's agent oversees cooling and electrical systems.
– Emerald AI agents adjust power based on live grid demand.
Extreme resiliency and efficiency,… https://t.co/325tqyQKuoJensen Huang highlights a new era of autonomous infrastructure. Using NVIDIA DSX Max-Q, AI agents now run the factory: – Phaidra's agent oversees cooling and electrical systems.
– Emerald AI agents adjust power based on live grid demand.
Extreme resiliency and efficiency, -
LennysData.com: Agentic Engineering for Data Platform Launch
By
–
Big thank you to @talraviv @clairevo @estelopez_ and Ben Shih for helping me beta-test and refine http://
LennysData.com (which I proudly “agentically engineered” myself ) -
KARL: Knowledge Agents via Reinforcement Learning Research
By
–
"You can think of this as RAG — but like RAG plus plus plus plus plus plus, all the way up to 200 vector database calls." — @jefrankle
, Chief AI Scientist at Databricks. We published new research on KARL — Knowledge Agents via Reinforcement Learning — a RAG agent trained