AI Coding Live News

YC Labs, a research project at Y Combinator, and multi-agent environment startup Olam Labs published "Agents in Games" on Oct 29. Fifty GPT-6.1 Sol agents told to "build a thriving civilization" in survival Minecraft spent three hours huddled in shelters rationing food, sending about 5,900 messages and inventing ration ledgers, permits and job titles. In nine Bed Wars games, 10 GPT-6.1 Sol agents beat 10 Claude Opus 5.5 agents 6–3; in seven 50-vs-50 Mindustry matches lasting about 9.5 hours, GPT won 6–1. The leader was almost always whichever agent was listed first in the directory, and stuck GPT agents fetched the game's public source code on their own. The Minecraft harness is open source. These are one lab's experimental observations, not a standardized benchmark.

Zoom

On Oct 28 xAI announced on the official @grok account that all API customers can now add a guidelines array to text requests and set the nsfw, violence and profanity categories to decline, so Grok tightens that content while it writes rather than through after-the-fact filtering. The field works on the Responses API, Chat Completions and the xAI Python SDK; multi-agent models and deferred completions are not supported, and invalid values are rejected with 400 or 422 before the model runs.

Zoom

On 2026-10-22, Anthropic's Applied AI team published "Building agents: When to fold subagents into your main agent loop" on claude.com, along with an open-source reference implementation, anthropics/single-agent-loop-kit. The article says that across several enterprise deployments, a single agent with skills frequently beat both the one-big-prompt design and the subagent design on quality, often at lower cost and latency per task; one customer that swapped its multi-agent harness for a single loop saw roughly 2x faster runtime. Anthropic also says it cut Claude Code's system prompt for more advanced models by over 80%. Subagents still make sense in three cases: measurable context bloat, parallel tasks with disposable context, or trouble selecting tools and conflicting instructions. The piece has drawn wide discussion on X over the past two days.

Zoom

Researchers introduced OverSeer in arXiv:2610.28554, pioneering a new class of multi-agent world models operating beyond the field of view of individual agents. Standard video world models collapse in multi-agent environments because individual agents operate under partial observability; naive frameworks feeding external global map inputs suffer severe spatial inconsistency and state hallucination. OverSeer resolves this by jointly predicting auxiliary spatial representation streams alongside the agents' first-person camera streams. In comprehensive Minecraft benchmarks, OverSeer achieves decisive state-of-the-art results across physical grounding, long-horizon spatial memory, and cross-view consistency.

Zoom

Artificial Analysis released an independent evaluation of Cognition's new coding agent, Devin Fusion. Moving beyond single-model execution, Devin Fusion employs an asymmetric multi-agent architecture pairing Claude Opus 5.5 (max) as the lead reasoning orchestrator alongside GPT-6 Luna (high) as a high-speed parallel worker. In the standardized Coding Agent Index, Devin Fusion achieved an elite score of 68.1 (matching Muse Code at 68.0 and rivaling Claude Code Sonnet 5.5 at 68.4). Crucially, the hybrid architecture slashes average task expense from $13.04 down to $9.01 (a 31% cost reduction) while shortening execution runtime from 65 to 55 minutes.

Zoom

To combat goal drift in long-horizon agent tool-use and multi-agent systems, NUS, USTC, and Alibaba Qwen introduced Steering Agent (arXiv:2610.22316). Jointly trained via pairwise ranking and Group-Relative Policy Optimization (GRPO), the agent operates a low-cost fast path to detect drift and a slow path to generate grounded interventions. It raises Multi-Hop QA benchmark performance from 42.70 to 50.10 while cutting executor turns from 8.8 to 6.5 (-26.1%) without modifying the underlying executor models.

Zoom

Alibaba Taobao & Tmall Group detailed CommerceGen-Studio (arXiv:2610.20603, 52 HF upvotes), a production multi-agent generation system serving millions of images daily within seconds. Its 6B SKU-DiT trained via Reinforce-Distill-Reinforce (RDR) scores 9.43 (single-image) and 9.45 (multi-image) on CommerceGen-Bench, outperforming 20B open models and closed-source Nano Banana Pro. Meanwhile, Microsoft Research introduced Mine, Sample, Render (MSR, arXiv:2610.21185) for training-free test-time diversity alignment.

Zoom

OpenAI announced a major upgrade to GPT-6 Astra Pro in ChatGPT for Pro users, introducing faster response times, stronger plugin and tool orchestration, and native parallel multi-agent execution where multiple agents tackle different parts of a complex task simultaneously. Concurrently, OpenAI Developers showcased how internal engineers put always-on 'dots' agents on call to autonomously triage 3 AM production incidents, launch Codex debugging threads, and clear codebase maintenance backlogs.

Zoom

Training data analysis agents is bottlenecked by privacy restrictions and unverifiable open-ended insights. Researchers introduce Data World (arXiv:2610.21656), a multi-agent engine that specifies underlying causal world mechanisms before generating heterogeneous files and verifiable rubrics across 1,037 disjoint worlds. Pure RL-zero training on Qwen3.6-35B-A3B yields DataWorld-35B-A3B, jumping +20.90 points on DABstep-Hard and +14.28 on DAComp. Alongside MRTPO (arXiv:2610.21439) for token-level multi-reward credit assignment, these works unlock scalable RL for data and SQL agents.

Zoom

Researchers from USC, University of Rochester, and NVIDIA introduce 3DCodeVerse, an open ecosystem empowering coding agents to synthesize editable and verifiable 3D worlds as code. Featuring the 3DCodeVerse-2M dataset of 2 million standalone executable 3D programs and a multi-agent DAG coding harness with automated spatial measurements and visual verification, 3DCodeVerse achieves a 90% human evaluation win rate against Claude Design (3D Object) when both run on Claude Opus 5.5.

Zoom

When model capabilities exceed human expert evaluative bandwidth, achieving Recursive Self-Improvement (RSI) without external verifiers or compilers becomes the paramount frontier of AGI. While reinforcement learning with verifiable rewards (RLVR) thrives in formal code and math, open-ended research and complex reasoning tasks lack scalar ground truth, and solitary models evaluating their own outputs suffer from homogeneous confirmation bias. Sakana AI introduces MASS (arXiv:2610.12176, project: sakana.ai/mass), an RSI framework anchored in Multi-Agent Self-Supervision. Driven by a single base model (such as Qwen3.6-27B), MASS tasks the model with concurrently acting as task solver, workflow optimizer, and self-evaluator. The framework alternates between evolutionary multi-agent workflow discovery and supervised fine-tuning (SFT) over synthesized trajectories. Across two successive MASS self-evolution cycles, performance per output token increases by 1.2x to 1.6x across open-ended research suites (MLR-Bench, ScienceAgentBench, DSBench, AstaBench). Head-to-head win rates against the base model surge from 53.9% in round 1 to 69.9% in round 2, while a student model trained on multi-agent traces decisively beats a single-agent baseline trained on 1.4x more tokens.

Zoom