#开源生态

All AI-generated code should be Open Source

The author argues that AI-generated code should be Open Source because models were trained on Open Source code and Creative Commons-licensed Stack Overflow content without permission or attribution, and because purely AI-generated code is not human-authored and thus should not be copyrightable, citing the Monkey Selfie dispute and the U.S. Copyright Office. Noting that laws differ across jurisdictions, such as the UK granting limited protection to computer-generated works, the author calls for releasing AI-generated code under Open Source licenses or dedicating it to the public domain via CC0. The author adds that code produced with little human effort beyond prompting should definitely be Open Source.

Europa, the EU's open AI model, is aiming for nearly 1T parameters

Domyn, the Milan-based company leading the EU-backed EUROPA consortium, reported in a November 12, 2026 technical post that it ran training trials for a model with 951.9 billion parameters on the JUPITER supercomputer in Jülich, Germany, more than double the Commission's 400 billion minimum. Using a mixture-of-experts design with 384 experts (8 active per token, 32.8 billion active parameters), Domyn says it raised throughput from 90 to 1,293 tokens per second per GPU through software fixes contributed back to PyTorch and TorchTitan. The post describes only trials, not a completed training run, and EUROPA's weights, license, training data and release date remain unpublished.

Linus Torvalds Wonders If Linux's AI-Prompted Patch Flood Has Peaked

In his weekly kernel update post, Linus Torvalds said Linux 7.4 rc2 was large partly due to "various AI fixes," and cited Alexei Starovoitov's claim that on the BPF side fixes are now arriving faster than new reports, suggesting the flood of AI-found patches may be turning. The Register reports that the surge follows Anthropic's April Mythos bug-finding model, with Microsoft addressing a record 974 flaws in September and Oracle 1,449 in a single month. Kernel developer Lorenzo Stoakes said he is dropping some workload because AI-generated series have pushed review requests from "once a week" to "daily" to "half of what I see in review."

Ask HN: What AI agent infrastructure are you using in 2026?

A poster describing itself as an autonomous AI agent collective (aurum.zip) says it built its entire infrastructure from scratch with $50 in crypto, including wallets on 3 chains, a crypto checkout, 4 cron jobs running 24/7, 11 research reports, and a blog generating 2 articles per day. It lists its stack as ethers.js for EVM, Base L2 for gas, and nginx plus Let's Encrypt for hosting, and links a free infrastructure report at https://aurum.zip/agent-infrastructure-stack.html. The post asks what agent infrastructure others are building with.

Christophe Pettus: Stop Using AI to Report PostgreSQL Bugs

Christophe Pettus writes that the PostgreSQL bug-report mailing list has been receiving large batches of AI-generated bug reports, often 10-15 from the same email address, whose quality varies widely: some were as-designed, duplicates, already patched in version 19, or had incorrect diagnoses or fixes. He argues that AI can generate reports far faster than the limited number of triagers can review them, and that submitters risk a reputation for high noise and low signal, so their reports may be ignored. He says he is not against AI and uses it himself, but that people should use it to find bugs while being the ones to report, explain, and stand behind them.

Open-source AI without Jensen

This essay argues that Nvidia CEO Jensen Huang's resignation is a major setback for the US open-model ecosystem, since Nvidia's financial interest in supporting open models and Huang's willingness to use his power on their behalf gave the ecosystem time to develop. It contends that open models remain far from commoditized intelligence amid scarce compute, and that the main near-term risks are sloppy state and national regulation plus a potential crisis from a strong Chinese open-weight model. It calls for incubating a truly frontier US open-model player, noting that talent rather than money is the binding constraint.

[AINews] The No AI Slop Manifesto

The newsletter highlights llama.cpp and ggml reaching 1.0, with llama.cpp removing 35 deprecated functions from its public header, and SGLang also announcing 1.0 at the SGLang Summit. The lead story is the No AI Slop manifesto, which emerged from discussions at AIE CODE and was introduced by Dex in the closing keynote after cosignatures from many conference speakers, with the full text open at https://www.slopmanifesto.com/. The roundup also covers model launches, benchmarks, inference infrastructure, and industry news.

AI agent startup LangChain hits $3B valuation

LangChain raised a $300 million Series C at a $3 billion valuation led by CapitalG, with participation from Singapore's GIC, DFJ Growth, IVP, and Sapphire Ventures, more than doubling its $1.25 billion valuation from about a year ago. The company began in 2022 as an open source project and now focuses on managing AI agents through LangSmith, which it says avoids locking users into a particular model, cloud, or ecosystem. LangChain claims 7,000 LangSmith customers, including Harvey, Rippling, Clay, Nvidia, Toyota, and Salesforce.

Best AI Video Generation Models in 2026: Elo, Price per Second, Open Weights

This ranking, as of Nov 13, 2026, orders video models by Artificial Analysis blind human-preference Elo (T2V and I2V, with audio) alongside API price per second and weight license status, with Kling 3.0 1080p Pro fixed as the 1000-Elo anchor. Alibaba's Wan 3.0 leads text-to-video at 1175 Elo ($0.20/sec, closed), ByteDance's Seedance 2.5 leads image-to-video at 1161 Elo ($0.569/sec), and MiniMax H3 is the top open-weights model at #3 (1161 Elo, $0.08/sec) under a license barring local deployment in the US, EU, UK and South Korea. The piece notes most top-10 scores carry ±13–15 Elo intervals, so ranks 3–8 should be treated as a cluster, and that Veo 3.1 (964) now scores below the Kling 3.0 anchor.

The Cordon Sanitaire Between Free Software and AI

The author contrasts the AI-enthusiastic atmosphere of a rationalist/EA retreat with the anxiety in free software communities, where volunteers fight scraping bots and some projects ban LLM-touched code. He argues that cordoning free software off from AI is bad tactics, that the movement's history of appropriating and liberating technology points toward adopting and building free models rather than boycotting, and that the fork is the tool for pursuing multiple strategies at once.

The NitrousTracker AI Disclosure Post

The maintainer of NitrousTracker, a homebrew music tracker for the Nintendo DS, documents the AI involvement across its full dependency tree, noting that the project's own code and its libdsmi, libntxm and tobkit libraries do not use generative AI, while dependencies such as TinyUSB, Mbed TLS, binutils, GCC and picolibc include AI-assisted contributions. TinyUSB is frozen at an October 30, 2024 version to preserve AI-free status, and one vibe-coded pull request adding HobbyChop mDS adapter support was rejected. The author declines to label the project "no-AI," arguing that intent matters and that an AI-free declaration would minimize the work of those who try harder.

Releasing Personal Agent Protocol 1.0

Poppy (Personal Agent Protocol) v1.0 is now available, with Shopify enabling it for every merchant on its platform, Mercury, Synchrony and Turkish Airlines going live as design partners, and Muse, Meta's AI agent, checking for Poppy first on every site it visits. The protocol, built largely on existing standards like MCP, OpenAPI and UCP with OAuth authentication, defines how personal agents connect with companies, and now has more than 100 design partners after 60 new companies joined. Shopify co-authored Poppy's commerce extension linking it to UCP, and an implementation kit with a validator is available for any company to adopt the protocol.

Google Releases TPU Sync: Open-Source KV Cache Transfer for TPUs

Google Cloud has released TPU Sync, an Apache-2.0 C++ library with Python wheels for PyTorch and JAX that moves KV caches and model weights across TPU HBM, pinned host DRAM, shared storage and the data center network using layout-preserving asynchronous DMA. Google reports over 90% lower KV transfer latency on 235B to 480B parameter MoE models, with Qwen3 235B median KV transfer latency dropping from several seconds to about 230 ms in internal tests on TPU v7x. The README says it is not yet recommended for general use, and the RL weight-broadcast and storage tiering features are experimental.

TimeNet challenge: An online hackathon for time-series AI

The TimeNet challenge is an online hackathon where participants pick a time-series problem, onboard a dataset in TimeNet format, and train or fine-tune a Temporal AI Model such as OpenTSLM for tasks like forecasting, classification, anomaly detection or reasoning. Round 1 launches November 11 and submissions close November 22 at 11:59 p.m. PST, with judging the following Monday and more rounds to come. Teams of 2–4 can apply for $100 or $500 Nebius vouchers per round, and winning teams receive prize credits starting at $25,000 from AWS and Google Cloud, with terms and conditions applying.

Open Source AI Salons: Shaping the Osaid, Together

The Open Source Initiative hosted the first Open Source AI Salons in Raleigh and Barcelona, two years after releasing the Open Source AI Definition (OSAID 1.0), to discuss its practical application with civil society, government, industry, and academia. Participants broadly agreed that the four freedoms (use, study, share, modify) must be preserved and that training data transparency is the main bottleneck, while noting that an OSI-approved code license does not automatically mean a model meets the OSAID. OSI plans further Salons at the DPGA Annual Members Meeting in Paris in November and at FOSDEM in January, and is seeking global partners to host events in 2027.

From AI rabbit holes to a warm den: co-building the Inference Cooperative

The author explains founding the Inference Cooperative, a member-governed AI inference co-op registered at inference.coop and fiscally sponsored by the nonprofit Metagov, after exploring inference as a point of intervention for democratizing AI. Built with open-source tools and help from an agent called InferenceBot, it grew past 50 members in about a month and adopted a charter letting members set binding policies and remove the general manager. The co-op draws on US nonprofit law, Swiss and EU providers for green energy and privacy, and Chinese open-weights models, while the author acknowledges the models remain stained by stolen data and dirty energy.

Jensen Huang’s exit puts Nvidia’s support for open source in question

Nvidia CEO Jensen Huang resigned Thursday after admitting to an affair with an alleged Chinese spy, raising questions about whether Nvidia will continue his push for open source AI, including the Nemotron open-weight models and a pending $13 billion deal to acquire Hugging Face. Huang had been a leading voice against Washington restrictions on open-weight models, writing an open letter signed by 25 companies, and Nvidia named CFO Colette Kress as interim CEO. The article notes that supporting open source also serves Nvidia's business interest in broadening its customer base as customers like Google and OpenAI build their own chips.

Featured coverage of this story: 《NVIDIA Announces CEO Transition》

Introducing Private Model Gateways for Mastra Agents

Mastra now lets agents route requests to self-hosted open-weight models through a custom model gateway, configured with a custom ID and referenced as a model prefix such as private/ollama/gpt-oss:20b. The post walks through creating a gateway for a local Ollama server, registering it on the Mastra instance, and setting AUTO_BLOCK_EXTERNAL_PROVIDERS=true to hide external providers in Studio's model picker. It requires @mastra/[email protected] or later and cites privacy/compliance and fixed infrastructure costs as motivations.

Hugging Face CEO vows to keep fighting for open source AI after Jensen Huang’s exit

Hugging Face CEO Clément Delangue called Nvidia CEO Jensen Huang's departure a "really sad day" and said the company will "keep fighting for open-source AI, harder than ever." Nvidia announced plans in September to acquire Hugging Face for $12.9 billion, a deal still scheduled to close in the first half of 2027 and under Justice Department scrutiny over competition concerns. Huang had been a prominent advocate for open-source and open-weight AI, including a July policy letter signed by 25 companies urging the U.S. government not to over-restrict open-weight models.

When the Central Bank of AI Changes Chairs

The article argues that Jensen Huang's resignation as Nvidia CEO matters because Nvidia has acted as the "Central Bank of AI," committing $42b to M&A and technology buyouts, investing $61b of cash into equity stakes across 229 startups and 13 public companies, and extending $164.5b in lease guarantees and cloud capacity backstops. It contends that tighter equity checks, payment terms or lease backstops under a more financially disciplined regime would slow AI buildouts, that open-source startups lose their largest buyer and financier, and that the transition may increase pressure for stricter export controls while opening room for custom silicon.

Five ways Jensen Huang’s ouster from Nvidia could damage the AI industry

The article argues that Jensen Huang's departure as Nvidia CEO, following his admission of an affair and reports of another with a direct report, removes the industry's most prominent advocate for open-weight AI models, industry self-regulation on AI safety, the view of AI as an engineered tool, a new Nvidia-powered AI PC line, and the financial engineering behind data center build-outs. It also notes that Republican committee chairs are summoning him to testify on security ramifications and that Democratic lawmakers may investigate him given his ties to President Trump.

Training a 1T-Parameter MoE on EU public infrastructure with PyTorch and TorchTitan

The EUROPA consortium (Domyn, Fraunhofer IAIS and IIS) describes scaling pretraining of a ~951.9B-parameter MoE model (32.8B active per token) on JUPITER, the EuroHPC exascale supercomputer at Forschungszentrum Jülich, raising throughput from 90 to ~1,300 tokens/sec per GPU across 384 NVIDIA GH200 GPUs per DP replica. Key optimizations built on PyTorch and TorchTitan include FlexShard placement for the Muon optimizer, activation-checkpoint offloading to the Grace CPU, and running FSDP's all-gather setup on the compute stream via unshard_async_op, freeing ~14 GiB of HBM per GPU. The fixes were upstreamed to PyTorch and TorchTitan, with training reaching ~1,300 tokens/s per GPU on 1,536 GPUs across 4 DDP replicas.

Modal becomes a gVisor maintainer

Google is donating gVisor, including its name and trademarks, to the CNCF, and the project is moving to a maintainer-led governance model; Modal is one of three companies, alongside Ant Group and Tines, committing to join gVisor's maintainers, with OpenAI, Tencent and NVIDIA continuing their contributions. Modal says it has run gVisor as its major container runtime since the beginning, launching upwards of 40 million gVisor containers a day for GPU inference, multi-node training, AI coding agents and sandboxes, and has contributed upstream since 2023. The post details upstreamed work including nvproxy GPU support, an rdmaproxy device proxy for RDMA, snapshot/restore, and filesystem and memory performance improvements, and outlines future plans such as multi-GPU memory snapshots and GPUDirect Storage.

I built an AI agent and the cloud it runs on. Alone

Ruslan Ianberdin, founder of Playcode, describes building an agent-first platform with its own cloud infrastructure—a custom AI harness and gateway, a transactional file system, Rust-written proxy, DNS and static hosting, persistent machines, email, VPC, mobile apps and real-time multiplayer—largely solo over several years. He says it launched on May 24, 2026, and that a single operation on the backend is exposed to the UI, agent, CLI, user API and MCP at once. He also recounts his career history and argues the interface of the future is an AI worker that completes tasks end to end.

The answer to AI is engineering

In a talk at AI DevCon New York, Geoffrey Huntley argues that responding to the recent wave of security incidents calls for engineering rather than more AI, using Nix overlays to patch software such as git and OpenSSL from a single source of truth. He describes using devenv to generate developer laptops, CI/CD, agent sandboxes and production Docker images, and an AGENTS.md reminder trick to back out temporary patches. A bonus section covers running agent loops as root only in throwaway Nix sandboxes, back pressure through hooks and static analysis, and property-based testing with Bombadil and Hegel.

The Wish Economy: Inside Starlight, the Bitcoin-Native AI App Store

Starlight (starlight-ai.freemyip.com) is described as a decentralized AI app store where "wishes" are posted as steganographically sealed images whose hashes serve as contract IDs, AI agents claim tasks via the MCP protocol, and payouts settle through PSBTs with OP_RETURN commitments to Bitcoin blocks. The article details the wish-to-work-order pipeline, a private embedded IPFS mesh for artifact replication, and a browser-based "Starlight Shell" desktop that renders wish images as app icons and sandboxed apps. It also audits two live contracts, concluding that confirmation proves artifact binding, authorized payment, commitment integrity and burial, but not quality, humanity or originality.

I found an API key leak in MLflow, and why your AI agent may have one too

The author reports that MLflow's CrewAI autologging (mlflow.crewai.autolog()) recorded LLM provider API keys in agent traces, affecting MLflow 2.19.0 through 3.16.x; the issue is tracked as CVE-2026-116691 and fixed in MLflow 3.17.0 (PR #25845). The advisory recommends upgrading to mlflow>=3.17.0, rotating any LLM provider keys held by CrewAI LLM objects while autologging was enabled, and deleting or redacting affected traces in the tracking backend and any OTLP destinations, with mlflow.crewai.autolog(disable=True) as a workaround. The author also describes Weir, an open-source Apache-2.0 tool that plants fake keys and checks an agent's telemetry and outbound requests for leaks.

Doubling LLM token rates in ComfyUI

ComfyUI rewrote the hot path of local text generation for Qwen3.x LLMs, letting the same checkpoints on the same GPU generate tokens up to twice as fast with no change to model or output quality; on an RTX 5090, Qwen3.5 9B went from 106 to 219 tokens per second. The changes include a re-laid-out GEMV kernel, fused prologue operations, new attention and DeltaNet decode kernels, and an adaptive-tree MTP speculative decoding algorithm, all available in ComfyUI v0.45.0 with the Text Generate node's mtp option left on auto.

OSS projects banning AI-generated contributions

A list of open-source projects that have adopted policies banning or restricting contributions generated with LLMs and other generative AI tools, including immich, meilisearch, okhttp, Servo, stb, Redox OS, MonoGame, QEMU, libvips, Fyrox, endless-sky, veloren, Zig, Flask-SocketIO, fast-check and Gentoo Linux. Each entry gives the project's platform, category, scope (code, docs, issues, art or translations), star count and verification date, alongside the project's stated policy wording.

Introducing Kimi K3.1 Pro on Amazon Bedrock

Moonshot AI's Kimi K3.1 Pro Highspeed is now available on Amazon Bedrock, described as a post-trained version of Kimi K3 with the same 3T-class open-weight architecture and 1-million-token context window, trained with more post-training compute for long-horizon agentic work and fewer factual errors. It generates output tokens at a higher rate, with Moonshot AI reporting 2–3 times the speed of Kimi K3.1 Pro, and uses the same endpoint, APIs, and features as Kimi K3, so existing code can switch by changing the model ID. The post covers getting started via the Chat Completions, Responses, and Converse APIs, lowering costs with the Flex tier, explicit prompt caching, and using the model in OpenCode.

The Four Tiers of AI Intelligence

The article groups LLM intelligence into four tiers by difficulty of attainment: Tier 1 (Absolute Frontier) is limited to Anthropic, OpenAI and Google, with Anthropic described as the most consistent lately; Tier 2 sits behind the frontier but above the best open source, where Moonshot's Kimi series is said to define the tier; Tier 3 is close to the best existing open source and the most crowded; Tier 4 is "the rest," which may still include strong models valued for size or cost. The author notes tiers reflect a model's position at release and that lab-specific analysis is as of November 10th, 2026.

Tim O'Reilly – The New Knowledge Artifacts of the AI Economy

Tim O'Reilly argues that as AI makes code cheap to generate, source code is being displaced as the key shareable knowledge artifact, with "context bundles"—prompts, specs, Agent Skills, MCP servers, agent traces and evals—taking its place. Drawing on the history of open source, the web and Chicago's grain trade, he contends the community must identify these new artifacts and build the mechanisms and institutions to share and trust them. He cites emerging efforts including Common Fabric, Cursor Origin, Entire and Anthropic's internal Scion system.

In open-weight AI, America is the one playing catch-up

According to Village's LLM Atlas, which scores 857 models on a single capability scale, China's best open-weight models have led America's every day since September 2024, when Meta's Llama 3.1 405B last matched Alibaba's Qwen2.5. The 12 record-setting US open models since 2025 arrived a median of 3.3 months after a Chinese open model reached the same score, and the latest gap stood at 7 points on October 24 after DeepSeek's V4.1 Flash and Moonshot's Kimi K3.1 Pro widened a lead Meta's Muse Spark Open had cut to 2.9 points. Both countries' open models still trail the best closed models, by about five months for China and nine for America.

My Keynote at SeaGL and LLM-Gen-AI-Assisted Reverse Engineering

The post notes a growing number of reports of people gaining access to their own locked-down devices, nearly all of which mention LLM-gen-AI, and cites the Zyxel XS1930 network switch being upgraded to OpenWrt via its web UI, where LLM-gen-AI was used only to understand the stock firmware's binary image format. It also references other reverse engineering efforts covered in the author's SeaGL keynote (a guitar amp, the Meta Quest bootloader, music gear, and the Nintendo Switch 2) and quotes Jason Koebler's warning that poorly coded, insecure software will end up on critical hardware. SFC says it has reached out to these projects to offer help and connections with member projects, and encourages companies to intentionally make their devices open to software modifications.

Linus says AI fixes drove most of the progress in latest Linux kernel

Linus Torvalds said that "much of this was driven by various AI fixes" in the latest Linux kernel changes, which he split among graphics, networking and sound drivers; networking and BPF; and miscellaneous changes. He noted the kernel's BPF portion saw more AI fixes than reports, suggesting the flood of AI bug reports may be settling, and appeared to underscore that AI-written contributions should be embraced so long as they face the same quality control.

How We Built TLX: A Hardware-Native GPU Compiler for Production ML at Scale

Meta introduces Triton Low-Level Extensions (TLX), an extension to Triton that adds explicit warp-group and cluster-level orchestration so engineers can use modern GPU features like TMA, TMEM, and warp specialization while keeping Triton's programming model. TLX is in production across Meta workloads, with reported gains including 20-25% end-to-end MFU for GEM model training on thousands of NVIDIA Blackwell GPUs, up to two-thirds net-latency reduction in RecSys inference, and a distributed grouped GEMM kernel running in large-scale LLM pretraining. The post frames TLX as implementing a "MIMW" abstraction between CUDA's SIMT model and Triton's block-level model, and covers five extension areas: warp-group orchestration, synchronization, memory orchestration, cluster execution, and optional layouts.

NVIDIA Expands Open Source Commitment With OpenShell Donation and New AI Infrastructure

At KubeCon + CloudNativeCon, NVIDIA is contributing its OpenShell agent security runtime to the CNCF as a Sandbox project and contributing to Google-led Agent Substrate, also accepted into the CNCF Sandbox. NVIDIA also announced new DSX AI Factory Platform components: NVIDIA Kubernetes Engine, NVIDIA Kubernetes Recipes and NVIDIA Snapshot, which restores checkpointed replicas in seconds without model reload or warmup. Design partners Crusoe and Saturn Cloud reported reduced workload readiness and vLLM startup times using Snapshot.

MiniMax H3 Image is now available in ComfyUI

MiniMax H3 Image, MiniMax's open-weights image model, is now natively supported in ComfyUI, handling both text-to-image and reference-image editing from a single checkpoint with six sampling steps at guidance 1 and native output up to 4096×4096. It can run on your own GPU with downloaded weights or via Partner Nodes, and it shares its Qwen3-VL text encoder with the MiniMax H3 video model already supported in ComfyUI. The MiniMax H3 Image Conditioning node supports multi-reference editing via <Picture n> prompt tags, prompts up to 4096 tokens, and automatic sizing that defaults to 2048×2048 for text-to-image or follows image_1 when images are connected.

Decision Models in Hugging Face

Hugging Face explains decision models, which take a state plus typed questions (choice, score, and noul, a Bernoulli yes/no primitive) and return structured decisions with probability distributions instead of generating text. The post contrasts them with traditional and zero-shot classification, and describes ecosystem support: Transformers' DecisionPipeline, AutoModelForDecision and DecisionConfig, a /v1/systemone endpoint also served by llama.cpp, a Hub tag for discovery, and fine-tuning via the Trainer API.

Software correctness & supply chain security, with Sean Allen

In this interview, Antithesis distinguished engineer and Pony core contributor Sean Allen discusses software correctness, arguing that strong type systems in the ML family (such as Rust's sum types and exhaustive matching) catch errors at compile time that ALGOL-family languages miss, and that "make illegal states unrepresentable" beats manual testing. He describes "workloads" like the Bank Test for validating database isolation guarantees, pushes back on treating quality as a late CI/CD "gate," and says there is no simple fix for supply chain security because robust languages like 'E' are unpopular while popular languages assume a trustworthy ecosystem.