Long-horizon tasks fail: Agents struggle with multi-step planning
rohanpaul_ai · x · 2026-08-29
Citing the Long-Horizon-Terminal-Bench paper, the post highlights that frontier agents struggle badly on long tasks, with an average pass rate of 6.4% across 17 models. 79% of failures are due to timeouts. Even the best model only finishes 28.3% of tasks. Chamath Palihapitiya is quoted criticizing the current state of long-horizon agents.
Related event: Chamath Slams AI at Stanford: Long-Horizon Tasks Still a Joke(5 posts)→
More from Models
- NVIDIA releases open Nemotron 3.5 Super VL: 120B params, 12B active, for multimodal agents — IanAndrewsDC · 2026-10-16
- New inference engine 'e' hits 10,500 tok/s on a single DGX Station — brandon_galang · 2026-10-16
- NVIDIA's Nemotron 3.5 Super VL hits CoreWeave: 12B-active multimodal MoE at $0.25/M input tokens — wandb · 2026-10-16
- Opus 5 might be a very deep model or use looping, commentator speculates — scaling01 · 2026-10-16
- GPT-6.1 Sol audit finds 6% of arXiv papers had false main theorems, down to 1.5% recently — nitarshan · 2026-10-16
- Nvidia-backed post-training pushes Nemotron 3.5 VL past frontier on dermatology diagnosis — LoganGrasby · 2026-10-16