Context Rot: How Deep Agents Keeps Long-Running Agents Focused
Context rot degrades long-running agents as tool output piles up. How LangChain's Deep Agents offloads, summarizes and isolates context, and what to tune.
AI Agents Context Reset: Fork-and-Flush Lifts Scores 66%
A context reset helps long-running AI agents escape idea basins. Fork-and-Flush beat single runs by 66% on equal compute. What it means for agent design.
Your AI Agent Is Silently Ignoring Your Instructions
When AI systems compress long conversations, they drop most user-defined behavioral rules. A Penn State study found only 17% of session constraints survive — and agents proceed without warning.
AI Agents Win by Throwing Away the Chat Log
Researchers replaced growing agent transcripts with a 5-layer structured memory system and cut token usage 90x while doubling win rates. The implication for agent architecture is significant.
Agent Memory: The Format Is Markdown — But the Problem Isn't
Andrej Karpathy, Google, and Garry Tan all independently chose Markdown for agent memory in 2026. They agree on the format and disagree on the problem it solves — and that distinction matters for how you build.
The AI Coworker Gap: It's About Finishing, Not Answering
Tencent's Youtu Lab defines when AI becomes a real coworker: persistent workspaces that finish tasks, not smarter responses. Most systems aren't there yet.
NotebookLM Now Has a Cloud Computer Per Notebook — and Agents That Find Sources for You
Google's NotebookLM update gives each notebook its own cloud computer for autonomous code execution, introduces agent-based multi-step research, and achieves 65% improvement over its previous version in internal testing.
ChatGPT Now Builds a Narrative Profile of You — Here's What It Knows
ChatGPT's new memory system creates prose dossiers covering your work, hobbies, and travel. Fact retrieval accuracy doubled to 82.8%. What this means for AI context.
OpenAI Codex Chronicle: Your Screen History Becomes Its Long-Term Memory
Codex's Chronicle feature watches your screen in the background, summarizes activity into persistent local memories, and uses that context to reduce cold-start friction — with real privacy tradeoffs.
Every Reply Is a Training Signal: What OpenClaw-RL Gets Right About Agent Learning
Princeton's OpenClaw-RL framework turns normal user interactions into continuous RL training data — no labeled datasets, no separate eval loop. Personalization scores jumped from 0.17 to 0.76 in 8 steps.