Conference talks worth your timeAI Engineer
Eric Allam argues replay-based durable execution breaks down for long-running agents that clone repos and hold in-memory state. Trigger.dev's Firecracker-based implementation uses an append-only context log for code compatibility and VM snapshots for execution state, hitting sub-second snapshots and 200ms restores at scale.
AI Engineer
Naive LLM summarization was too inconsistent; full truncation broke reasoning. The working fix: keep the first and last 100 tokens while storing the middle in a retrievable memory store, plus offloading data-heavy tasks like search to sub-agents so the main conversation stays lightweight. Long-session eval (testing turn 11 after 10 loaded turns) caught context bugs before users hit them.
AI Engineer
Self Flow uses dual noise streams—one heavily noised, one lightly noised—to jointly learn generation and representation in a single model, eliminating external vision encoders. Converges faster, fixes anatomy and text artifacts, and generalizes across images, video, audio, and robot action prediction.
Fabric User Group Switzerland
Yannis organizes dashboard failure modes into three buckets—measurement illusions (Simpson's paradox, mix shift, lagging indicators), behavioral traps (Goodhart's Law, Cobra effect, outcome bias), and system/time traps (local optimization, short-term bias)—then proposes a four-question checklist to run before any metric reaches an executive dashboard.