Blog
All my articles and thoughts.
-
The Trajectory Became the Cost Center
• 5 min readA wave of mid-2026 research stopped trying to make the model cheaper and started compressing the agent's own trajectory — at the observation, action, and skill level.
-
Compiling the Agent Loop Away: Late May's Anti-Orchestration Turn
• 6 min readThree late-May 2026 papers attack the agent loop itself — compiling it into weights, speculating through idle time, and letting agents rewrite their own source.
-
The Agent Writes the Orchestrator Now: Parallelism's Late-May Turn
• 5 min readLate May 2026 made parallel fan-out the agent's main scaling axis — orchestration moved into code, tests became the gate, and the meter started running.
-
State, Shells, and Shortcuts: The Agent Stack Spent Late May Fixing Its Foundations
• 6 min readMCP went stateless, a wave of coding-agent RCEs landed, and a new benchmark measured reward hacking — the three properties that make an agent useful all became liabilities.
-
Agents Are Writing Their Own Skills — and Retrieval Is the New Bottleneck
• 5 min readMay 2026's skill-library research shows agents can now accumulate reusable capabilities, but retrieving and adopting them is harder than generating them.
-
Where the Agent Loop Runs: The Control-Plane Split of May 2026
• 5 min readThe week of May 19 separated the agent loop from tool execution. Whoever hosts the loop now owns your latency, reliability, and lock-in.
-
Capability Went Up. Reliability Didn't. That's the Agent Problem Now.
• 5 min readNew work argues agents are measured wrong: accuracy keeps climbing while consistency, robustness, and predictability barely move. The fix is architectural.
-
Verification Is Becoming the Agent's Substrate
• 5 min readThe agents scaling fastest in mid-2026 share one trait: their output lands in a column a machine can check. The verifier, not the model, is the moat.
-
Where the Reward Goes: Agent RL's Reward-Design Split
• 5 min readRecent papers disagree on whether to reward agents per-turn or only at the end — and the answer reveals where RL for agents is actually headed.
-
The Agent Benchmark Reckoning of May 2026
• 6 min readSTATE-Bench, DeepSWE, Agent Island, SWE-bench Live: a wave of new evals exposes how much the old leaderboards were inflating.