Mike Saleme — 2026-04-20 — views my own This week OpenAI released GPT-5.4-Cyber, positioned as the...
The Mythos vs GPT-5.4-Cyber debate is missing the benchmark
Related
I kept getting lost in internship meetings, so I built an open-source AI overlay in Electron
During meetings with my internship team I kept getting lost. Someone would mention jargon I didn't...
What Actually Happens When an AI Agent Loops: The ReAct Pattern, Context Window Poisoning, and How to Break the Failure Spiral
Why autonomous AI agents get stuck in repetitive failure loops, how autoregressive attention and context poisoning reinforce bad actions, and how to build deterministic circuit bre...
My Agent Said We Placed. The Leaderboard Said Otherwise.
What running an AI agent taught me about the difference between "trust me" and "check it" — my entry...