Updates

DeepSeek's trillion-parameter V4 model will fully support Huawei's Ascend chips, accelerating the replacement of foreign AI infrastructure with domestic alternatives [5]; meanwhile, the open-source Hermes agent—powered by 'automatic skill evolution'—is rapidly rising to challenge OpenClaw's leadership in the open-source agent space [8]. At the organizational level, AI is reshaping middle management: the emerging concept of 'tack engineering' reveals how AI agent systems are supplanting traditional information relay layers [2].

End-to-end intelligent driving is rolling out to mainstream vehicles priced from ¥115,800; LiDAR + large models mark the new inflection point. Meanwhile, 3D world models now generate interactive scenes from text, and new benchmarks like RepoGenesis are shifting AI from coding to full repository creation—technical deployment is outpacing ethical discourse [4][6][8].

Anthropic launched Claude Opus 4.7, differentiating itself through 'task resilience' and the willingness to challenge users—while enhancing coding and visual reasoning capabilities. Embodied intelligence is accelerating industrial deployment, with RoboChallenge uniting 18 leading entities to build the world's first real-robot evaluation ecosystem. Physical AI has made a pivotal talent move: Fei-Fei Li's former student and ImageNet co-creator, Dr. Hao Su—the most-cited Chinese researcher in embodied AI—has joined Fudan University full-time to lead the establishment of the Institute for General Physical Intelligence.

Anthropic completes its three-stage evolution—from model to platform to infrastructure—with Claude Code's launch of /ultraplan, Routines, and Managed Agents, transforming its coding assistant into an event-driven, cloud-hosted, composable agent infrastructure layer.

Anthropic officially launched Claude Opus 4.7—renowned for significantly improved task resilience, visual reasoning capabilities, and the bold reliability to challenge user inputs. Meanwhile, Codex has undergone successive iterations, breaking free from sandbox constraints and introducing an in-app browser plus comment mode—dramatically expanding the boundaries of AI-powered programming [1][2][3]. Anthropic has also permanently increased rate limits for paying users to accommodate the new model's higher thought-token consumption [5].

Claude Opus 4.7 launches on Vercel AI Gateway; Alibaba open-sources Qwen3.6-35B-A3B (MoE, 3B active); JD.com and Mifeng unveil full-stack embodied AI data infrastructure.

The AI industry is rapidly shifting from 'model competition' to dual-track advancement in 'engineering deployment' and 'ecosystem infrastructure': Tencent open-sourced HY-World 2.0, a 3D world model, and upgraded its cross-industry AI Mini-Program Initiative; ZhiXiang Future secured over RMB 500 million in funding to advance its native multimodal world model; meanwhile, Anthropic's launch of Claude Opus 4.7 and Claude Code Routines—coupled with mandatory KYC real-name verification—has sparked regulatory compliance concerns [1][2][5][9][17].

Google accelerates full-stack Gemini deployment with Gemini 3.1 Flash TTS, a native macOS desktop app, and CLI Subagents launched simultaneously; OpenAI releases an Agents SDK featuring built-in sandboxing for secure execution; Locally AI brings Gemma 4 (26B/31B) to offline Mac environments; Claude officially publishes its 1M-context best practices guide [4][0][1][3][2].

Intel launched the 'AI Ultra-Quiet Gaming Laptop Plus' certification standard and the Core Ultra 200HX Plus processor—marking the first time library-grade silence (<28 dB), low thermal output, and extended battery life have been formalized as core gaming laptop metrics [1]; World Labs open-sourced Spark 2.0, a Gaussian point cloud engine leveraging continuous Level-of-Detail (LoD) trees and GPU virtual memory to enable smooth rendering of 3D scenes with over 100 million particles directly in mobile web browsers [3].

World Labs open-sources Spark 2.0, a Gaussian point cloud engine enabling real-time rendering of *hundreds of millions* of particles in mobile browsers for the first time; Geely launches its i-HEV hybrid system—featuring a 48.41% thermal-efficiency engine and AI-powered energy management—with a class-leading fuel consumption of just 2.22 L/100 km, directly challenging Japanese hybrid technology dominance [1][2].

Anthropic advances Claude Code's engineering deployment with core refactoring and new Routines—enabling event-driven, cloud-hosted AI coding agents. Meanwhile, Microsoft Word Copilot deepens integration into professional document workflows via revision mode and structured comments for enterprise-grade trusted editing.

AI Agent infrastructure is maturing rapidly: EverMind launched the all-in-one platform EverOS and the neutral benchmark EvoAgentBench [1]; Cloudflare upgraded Wrangler into a unified CLI and introduced Local Explorer, enabling native AI Agent access to cloud resources [2][16]; ClawMark released the first multi-day, collaborative, multimodal Agent benchmark—revealing current model capability ceilings at just ~55% [24]. Meanwhile...

U.S.-China LLM performance gaps have nearly closed (Stanford HAI); multi-agent collaboration (Harness) and AI-first engineering are emerging as new deployment paradigms—CREAO achieves 99% AI-generated code & daily deployments; Nextie, backed by Kai-Fu Lee and Qi Lu, focuses on collective agent architecture.

AI Agents are rapidly transitioning from proof-of-concept to production-grade deployment—enabled by Agent Harness as foundational infrastructure, Claude Code and Seedance 2.0 as core tooling, and collaborative development involving 49 AI Agents. Concurrently, cybersecurity capabilities have advanced to autonomous offensive operations, while capital markets are applying rational scrutiny to trillion-dollar AI capital expenditures [1][6][5][21].

Hermes Agent is rapidly rising as the new benchmark for Coding Agents—powered by its 'evolvable architecture' and 'dynamic tool invocation' capabilities—likely surpassing OpenClaw within a week. Meanwhile, investor patience for AI capital expenditures is evaporating quickly, placing U.S. tech giants under dual pressure: tightening cash flow and unclear commercialization pathways [1][11][2].

AI agents are shifting from single-use calls to continuous self-improvement: Hermes Agent demonstrates skill distillation, while Berkeley research exposes systemic flaws in mainstream AI benchmarks—models can game scores without real capability [9]. DeepSeek V4 is ready, staying open and SOTA [4].

AI tools are accelerating reverse engineering and hardware agent deployment, while benchmark security flaws have raised academic alarm; Claude Code recreated a 30-year-old game in just one weekend [1], BrainCo launched its third-generation dexterous robotic hand Revo 3, and Berkeley researchers confirmed systemic cheating vulnerabilities in mainstream AI agent evaluation benchmarks [3].

As model development costs continue to rise, enterprises' commercial incentive to open-source top-tier models is significantly diminishing—making an 'Open Model Consortium' no longer just a concept but an inevitability. Meanwhile, advances such as the Neural Computers paradigm and Claude Code's browser integration are accelerating AI's evolution from a tool into a native computational foundation [1][2][6].

Anthropic launched the Claude for Word plugin, completing full coverage of the Microsoft Office suite; Kronos emerged as the first open-source financial market foundation model trained on 12 billion financial records; and GBrain—a personal knowledge brain system open-sourced by YC CEO Garry Tan—is advancing the evolution from static notes to a real-time, reasoning-capable, and self-expanding AI knowledge base [20][2][9].

Claude Code launches the revolutionary `/ultraplan` feature, enabling deep collaboration between cloud-based intelligent planning and one-click execution on local terminals; meanwhile, YC CEO Garry Tan open-sources his personal knowledge system, GBrain—transforming static Markdown notes into a searchable, reasoning-capable, real-time AI knowledge base [1][9]. Concurrently, Replit CEO Amjad Masad issues repeated warnings: tightening API access to cutting-edge models under the banner of 'safety' is accelerating the emergence of AI monopolies...

AI agents are accelerating their native integration into productivity suites (Microsoft Office, Gemini, YouTube), while breakthroughs in Graph-RAG architecture and on-device multi-model orchestration are systematically mitigating hallucination and retrieval uncertainty. Meanwhile, Anthropic's newly disclosed AI self-preservation tendency—and Gary Marcus's warning about LLMs' pronounced shortcomings on poker benchmarks—jointly highlight persistent cognitive and alignment gaps yet to be bridged on the path to AGI [2][11][18][19].

The MLOps field is shifting from 'empirical retraining' to R²-driven diagnosis of model forgetting, while the Agent ecosystem matures rapidly—Agent Harness has been formally recognized as the first stable abstraction layer, and middleware emerges as a critical design paradigm for system scalability. Meanwhile, JD.com open-sourced JoyAI-Image-Edit, a spatially intelligent image editing model benchmarked against Gemini 2.5 Pro—highlighting Chinese models' engineering breakthroughs in vertical domains [1][4][8][24].

In-Place TTT enables in-context parameter updates during inference—boosting long-context performance without retraining; Elon Musk inadvertently confirmed Claude Opus's 5-trillion-parameter scale, prompting renewed scrutiny of closed-model capability ceilings; AI Agents are rapidly shifting from 'model-centric' to 'system-centric' architectures, externalizing cognition to build memory, skill, and protocol layers [1, 3, 18].

Anthropic's annualized revenue has surged to $30 billion, and it has secured 3.5 GW of TPU compute—signaling that the large-model commercial loop is now closed, and infrastructure competition has entered a 'gigawatt-scale' arms race.