Saturday, May 23, 2026
6 stories covered by gentic.news intelligence
Claude Code /workflows replaces LLM orchestrator with code-based control flow, solving the token tax problem from multi-agent context buildup.
Cerebras reported 981 tokens/sec on the 1T-parameter Kimi K2.6 model, a 6.7× speedup over the next GPU cloud, validated by an independent third party.
Opus 4.8 spotted on Vertex AI. Unconfirmed but plausible given GPT-5.5 pressure. Watch for official confirmation.
Karpathy predicts neural networks will become the host OS, with CPUs as co-processors, rendering most classical app interfaces obsolete.
Paper shows 88.5% improvement by adapting runtime interface around frozen LLM. Harness generalizes across 18 backbones, challenging model-centric agent improvement.
Google paper shows 18% heart rate accuracy gain by personalizing wearable AI to individual users via lightweight embeddings.