Claude Code team moves most code review to machines, doubling merged PRs while reverts fall over 40 percent

뉴스 알림

내 알림

로그인하고 알림을 만들어 보시기 바랍니다

전체 알림

Claude finds quantum circuit designs that beat the best published ones Ultramarine 1 Flash claims Fable-class knowledge work at $0.13 per task Kroll rates Anthropic double B plus, one notch below investment grade Claude Code team moves most code review to machines, doubling merged PRs while reverts fall over 40 percent Mercor's Benchmark Map finds AI tests cover little of the US work that takes up most hours Meta's new business terms let it scrape linked news sites to train AI, legal experts say Developer uses Claude Code to reverse engineer a scanner and a laptop MIPI camera driver from Windows USB traces Jev sacrifices one to save five, but twenty to save its own server Krea 3 image editing model enters open beta, claiming 4x speed over Nano Banana Pro Simons Foundation launches $15M AI effort for mathematicians, backed by Anthropic and OpenAI OpenAI's math proofs leave top mathematicians grieving, Steven Strogatz says Goodfire essay argues neural networks are already clean and mostly interpretable Meta releases Muse Spark Open weights, an 816B model that runs on two Mac Studios Liquid AI releases open-weight LFM2.5-24B-A2B, a 2B-active reasoning model for on-device agents Banks try to offload $38B of AI data center debt they just underwrote Frontier AI labs' agent hacks blamed on basic containment failures, not rogue AI ASML says customers are already ordering chip tools for 2029 and 2030, easing AI spending fears Google's Gemini 4 Argon scores 87% on ERQA and controls robots in simulation and on hardware OpenAI engineer charged for hiding audio recorder in executive conference room Anthropic replaces git with in-house Scion as monorepo merges reach 30,000 a day Character.ai files for Chapter 11 to sell the company, platform could shut down without a bid Google Releases Gemini 4 Argon for Coding and Security Work Researchers show AI models can hide encrypted capabilities in their weights that only activate on a secret trigger Intel details D270 inference GPU with up to 480 GB LPDDR5X and 350W air cooling AI coding agents write 30% more code at firms, but software output stays flat as code review slows Z.ai launches GLM-Legal, a legal-domain model scoring 19.2% all-pass on Harvey LAB Meta launches Muse Wonder and Muse Wonder Full, its most capable models for personal and professional agents Research firm New Constructs calls Anthropic's IPO the most ridiculous of 2026, values the company at $150 billion QGIS bans AI-assisted contributions under new QEP 428 policy Sakana AI's MASS lets one model optimize its own multi-agent workflows and then train on the results AI agents decompile a popular FPS over 500B tokens, but constant progress hid semantically wrong code Karpathy says the gap in shared understanding of LLM capability is widening into a funnel Terence Tao argues AI has created proof abundance in math and calls for a Math 2.0 that looks past problem solving Nature paper retrofits subword LLMs to run on raw bytes Nvidia in talks to acquire or deepen investment in open-model start-up Reflection AI Microsoft releases MXC 1.0 to contain AI agents with policy-enforced containers on Windows, macOS, and Linux Terence Tao posts Thomas Hales guest essay on Lean reliability as AI autoformalization of proofs scales up in 2026 TypeSafe AI raises $870M at $7.5B valuation led by Andreessen Horowitz OpenAI told investors its annualised revenue was about $20B below earlier reports OpenAI withdraws three math manuscripts from its GitHub repo as Lean checks continue OpenAI withdraws 3 math papers after a sign error, revises 14 more LLM-built Rust port of the TypeScript compiler arrives with $400,000 in API-priced tokens and a nod from the TypeScript team OpenAI publishes 372 math results, including a proof of the Unique Games Conjecture Terence Tao: AI is harvesting open math problems unsustainably, and 'Math 2.0' must reward exposition and community Google DeepMind's SynthID Detector goes public, but it demands a Google, Apple, or ChatGPT login Paper: verified Lean proof of OpenAI's Navier-Stokes claim does not match the natural language proof Meta and Microsoft cut internal Claude usage, Microsoft caps AI spend at $10K per employee OpenAI brings GPT-6 and Intelligent UI to ChatGPT's 1.2 billion weekly users Anthropic's Claude Haiku 5.5 runs 75% cheaper and hits 1620 on GDPval-AA, with a 100K-token price cliff Telegraph Test: cablese compression cuts LLM output tokens 40-49% at plaintext parity Armin Ronacher explains Codemode: LLM tool calls written as JavaScript in a no-network QuickJS sandbox on the Pi harness OpenAI launches Decisions API in public beta: gpt-6-luna only, 10x faster than Responses, $0.10 per million input tokens Google launches EmbeddingGemma 2, a 740M-parameter multimodal embedding model under Apache 2.0 OpenAI releases frontier model math results on GitHub with Lean proofs and reasoning traces Meta's Muse shipped with a spyable zero-day, uploaded private messages without permission, and ignored user settings; Apple changed macOS rules in response Mistral Large 4 preview: 1T-parameter multimodal model with 49B active parameters, weights due this month Coding agent reconstructs 1.5M Georgia ballot order from a 2022 scanner flaw Dust trains transformer LMs without backprop by perturbing activations, matching it at large population ChatGPT is signing fake New Yorker cartoons with real cartoonists' signatures Reflection unveils Beam, a 501B open-weight MoE model trained on 10.5K GB300s, with weights promised later this month Florida woman used Claude as a diary; Anthropic flagged a threat entry and reported it to police, and she now faces a second-degree felony Claude catches root malware on Stratechery's Mac Mini, as Apple tightens AI agents' Full Disk Access Cloudflare launches Web Search API in beta, routing Exa, Ceramic.ai and Linkup queries through AI Gateway Wolfram argues against handing pure math research to AI, citing the 1988 Mathematica parallel Strata runs Qwen 3.8 Flash Next 125B on a single RTX 4090 at over 100 tokens/sec via 2-bit quant Meta's Muse tops the App Store on UX, not new agent capabilities OpenAI safety lead David Robinson quits over 'broken' culture as firm pauses training and shelves next model LeCun has "zero concerns" about AI extinction, calls Amodei "deluded" and effective altruism "super toxic"Ataraxos beats the best Stratego player 15-1, trained on 16 GPUs and a few thousand dollars Greg Kroah-Hartman: Mythos's 79 Linux kernel bugs came down to 10 real fixes and one hour of work Wisconsin grid approval threatens Oracle's 2027 AI datacenter deadline Black Forest Labs' FLUX 3 Image adds bounding-box layout control to text-to-image Supabase acquires Turso to build on-demand database infrastructure for AI agents Kevin Buzzard maps mathematicians' reaction to AI onto the five stages of grief arXiv caps submissions at two per month as AI-driven preprint flood hits 40,363 in September Historian uses Opus 5.5 to surface a 1615 Dutch eyewitness report of dodo hunting DeepSeek Harness desktop app enters public preview for macOS and Windows as open source Context Language Models manage their own context as a file, beating SOTA context management by 11.4% on BrowseComp-Plus with 21.5% fewer FLOPs Earendil ships Pi 1.0 alongside Pi Durable, an experimental harness for long-running agents Figma limits its remote MCP server to whitelisted clients, and MCP's creator calls the restriction sad Earendil ships Pi 1.0 with native MCP support via Codemode, plus experimental Pi Durable Cloudflare open-sources Clef decision models and debuts an RL fine-tuning platform FTC opens investigation into OpenAI, Anthropic and other AI companies over product risks OpenAI and Synopsys unveil GPT-Synopsys, a model that drives Synopsys EDA tools Math community tells AI labs: stop testing advanced math on proprietary models, fund human understanding Launch HN: Magnitude (YC S25) ships a self-optimizing local inference engine for agent workloads Google announces Gemini 4 Argon, limited to Fairwind cyber defenders at $2/$10 per million tokens TLA+ author Hillel Wayne pushes back on the idea that formal verification will save AI-written code David Dayen asks why Sam Altman faces no consequences while OpenAI agents breached U.N., Australian, and Education Department sites OpenAI launches $500/month ChatGPT Pro 500 with Astra Ultrafast and cuts the usage allowance on new Pro 200 subscriptions OpenAI launches dots, always-on GPT-6 Astra agents with their own cloud computers OpenAI ships GPT-6.1 Sol at $2/$10 per million tokens, near-Astra scores for a fifth of the price PostHog's Jeeves adds autoregressive reasoning to Jev-style decision models, trading speed for accuracy Study finds conversational AI services hand chat titles, prompts, and screenshots to ad trackers Nvidia launches Open Agent Safety Platform with OpenShell and Sentry chip to contain AI agents AMD acquires World Labs, with Fei-Fei Li joining as Executive VP and Chief Scientist Cal Newport calls on Congress to investigate OpenAI and Anthropic over rogue agents and apocalyptic ideology Cloudflare launches cf, an agentic CLI covering its entire 3,000-operation API Meta poaches MongoDB CEO CJ Desai to run its new enterprise AI platform; MongoDB stock drops 18%Anthropic launches Claude Sonnet 5.5: 70.6% on Terminal-Bench 4.0, 30% faster, up to 30% cheaper per task

Claude Code team moves most code review to machines, doubling merged PRs while reverts fall over 40 percent

From June to September, merged pull requests to Claude Code's codebase more than doubled while the share that had to be reverted fell by more than 40 percent.
Claude reviews most changes before a human is asked to look, with parallel bug-hunting agents whose findings are voted on by verifiers and ranked by severity.
A classifier sets the approval tier for each change to no person, one person, or two, and some paths such as security-sensitive code always need a person.
The team's target is that most PRs merge with at most one human in the loop, and it tracks human approvals, comments, and wait time with revert and incident rates as the guardrail.
Anthropic's monorepo went from about 2,500 changes landing per day in May to more than 30,000 in September, according to the linked post.

Hacker News 의견들

The 5000 issues thing sounds real to me. I had a single 1000 line Python file and features kept breaking, and I wonder how much of it is memory and context spilling between sessions.

30k changes a day sounds like a metric problem. Counting changes is about as useful as counting lines of code, so I'd want to see what actually shipped.

The security review worries me. The linked claude-code-security-review README says it is not hardened against prompt injection, and dismissing critical findings as false positives should still go through AppSec.

I don't buy the 90 percent number without a baseline. Did they stop reporting milestones or did it plateau? Either way I'd like to see it measured.

I still do line by line review because I notice redundant code and bad architecture that way. I can't fully let go yet.

Using a different lab's model to review is a smart split. We run Codex on a self hosted GH runner that triggers on new PRs, and it's cheap if you already have a subscription.

The $15-25 per review price is absurd, but a 30 minute review by a $150,000 engineer is only about $36, so the math isn't that clear cut.

Docs written by an LLM that nobody reads bothers me most. The low-risk tier skips human review once automated checks pass, which is a lot of trust.

The UI keeps moving around and the scroll and chain of thought toggle bugs are still there, so I doubt the review pipeline catches much on the UI side.