惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
Exploit-DB.com RSS Feed
Exploit-DB.com RSS Feed
The Last Watchdog
The Last Watchdog
K
KPMG report finds enterprise disconnect between AI and its ROI | CIO
T
Troy Hunt's Blog
L
LINUX DO - 最新话题
C
Check Point Blog
T
Threat Research - Cisco Blogs
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
罗磊的独立博客
V
Vulnerabilities – Threatpost
cs.CV updates on arXiv.org
cs.CV updates on arXiv.org
J
Java Code Geeks
Apple Machine Learning Research
Apple Machine Learning Research
大猫的无限游戏
大猫的无限游戏
S
Security @ Cisco Blogs
IT之家
IT之家
T
The Exploit Database - CXSecurity.com
The GitHub Blog
The GitHub Blog
D
Docker
Engineering at Meta
Engineering at Meta
AWS News Blog
AWS News Blog
S
Security Affairs
U
Unit 42
P
Palo Alto Networks Blog
V
Visual Studio Blog
Y
Y Combinator Blog
D
DataBreaches.Net
Forbes - Security
Forbes - Security
阮一峰的网络日志
阮一峰的网络日志
美团技术团队
Security Latest
Security Latest
aimingoo的专栏
aimingoo的专栏
Simon Willison's Weblog
Simon Willison's Weblog
A
Arctic Wolf
博客园_首页
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
H
Hacker News: Front Page
博客园 - 司徒正美
博客园 - Franky
宝玉的分享
宝玉的分享
TaoSecurity Blog
TaoSecurity Blog
Latest news
Latest news
Scott Helme
Scott Helme
MongoDB | Blog
MongoDB | Blog
量子位
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
C
Cisco Blogs
P
Privacy International News Feed
Application and Cybersecurity Blog
Application and Cybersecurity Blog

Hacker News

Introducing Claude Opus 4.7 Qwen Studio The Future of Everything is Lies, I Guess: Where Do We Go From Here? GitHub - SeanFDZ/macmind: Single-layer transformer in HyperTalk for the classic Macintosh Show HN: Agent-cache – Multi-tier LLM/tool/session caching for Valkey and Redis Moving a large-scale metrics pipeline from StatsD to OpenTelemetry / Prometheus GitHub - Nightmare-Eclipse/RedSun: The Red Sun vulnerability repository GitHub - SethPyle376/hiraeth: Local AWS emulator focused on fast integration testing, with SQS support, SQLite-backed state, and a debug-friendly web UI. GitHub - macOS26/Agent: Any AI, replaces Claude Code, Cursor, OpenClaw. Over 18 LLM providers (Claude, OpenAI, Gemini, Ollama, Zai, HF, Qwen) wired into a native Mac app that writes code, builds Xcode projects, bumps versions, manages git, automates Safari, use AppleScript, JS or Accessibility, extend Agent! w/ MCP Servers, run tasks from your iPhone via Messages. YouTube now lets you turn off Shorts I Made a Terminal Pager Burgers | マクドナルド公式 Commands — HackerNews CLI documentation ChatGPT for Excel PiCore - Raspberry Pi Port of Tiny Core Linux Live Nation illegally monopolized ticketing market, jury finds Google Broke Its Promise to Me. Now ICE Has My Data. Founding Engineer at Adaptional | Y Combinator CRISPR takes important step toward silencing Down syndrome’s extra chromosome GitHub - saffron-health/libretto: The AI toolkit for building reliable browser automations US v. Heppner (S.D.N.Y. 2026) no attorney-client privilege for AI chats [pdf] Unexpected €54k billing spike in 13 hours: Firebase browser key without API restrictions used for Gemini requests Retrofitting JIT Compilers into C Interpreters IPv6 – Google The Accursèd Alphabetical Clock Cybersecurity Looks Like Proof of Work Now Fragments: April 14 Cal.com Goes Closed Source: Why AI Security Is Forcing Our Decision | Cal.com - Scheduling Software for Online Bookings Laravel raised money and now injects ads directly into your agent When moving fast, talking is the first thing to break Too much Discussion of the XOR swap trick – Heather Cafe Introduction to Spherical Harmonics for Graphics Programmers The Grand Line Building a Z-Machine in the worst possible language High-Level Rust: Getting 80% of the Benefits with 20% of the Pain GitHub - duguyue100/midnight-captain: Inspired by Midnight Commander, tailored to my taste. How to build a `git diff` driver · Jamie Tanna | Software Engineer Center for Responsible, Decentralized Intelligence at Berkeley The Local Universe’s Expansion Rate Is Clearer Than Ever, but Still Doesn’t Add Up - A new synthesis of astronomical measurements confirms a persistent mismatch that could point to physics beyond current models The air throughout our homes is infused with microplastics. But there are things you can do to breathe less of them The disturbing white paper Red Hat is trying to erase from the internet – OSnews The Future of Everything is Lies, I Guess: Annoyances ‘Abhorrent’: the inside story of the Polymarket gamblers betting millions on war Productive procrastination — Max van IJsselmuiden maps, territory and LMs 447 Terabytes per Square Centimetre at Zero Retention Energy: Non-Volatile Memory at the Atomic Scale on Fluorographane Show HN: Pardonned.com – A searchable database of US Pardons 20 Years on AWS and Never Not My Job The Seasons are Wrong Artemis II crew splashes down near San Diego after historic moon mission We gave an AI a 3 year retail lease in SF and asked it to make a profit | Andon Labs How a dancer with ALS used brainwaves to perform live On filing the corners off my MacBooks Installing every* Firefox extension OpenClaw’s memory is unreliable, and you don’t know when it will break Steve Blank Nowhere Is Safe Chimpanzees in Uganda locked in vicious 'civil war', say researchers watgo - a WebAssembly Toolkit for Go linux/Documentation/process/coding-assistants.rst at master · torvalds/linux GitHub - callumlocke/json-formatter: Makes JSON easy to read. Founding Product Engineer at Bild AI | Y Combinator A compelling title that is cryptic enough to get you to take action on it GitHub - Keychron/Keychron-Keyboards-Hardware-Design: Industrial design files for Keychron keyboards and mice. 100+ models with CAD assets in STEP, DXF, DWG, and PDF. Source-available, with commercial use allowed for original compatible accessories within the license terms. [ANNOUNCE] WireGuardNT v0.11 and WireGuard for Windows v0.6 Released 1D-Chess Helium Is Hard to Replace Cooperative Vectors Introduction | Evolve Keeping a Postgres queue healthy — PlanetScale Our response to the Axios developer tool compromise Do Americans read print books, e-books or audiobooks more? The Zettelkasten Method in Obsidian: A Practical Setup Guide Artemis II Is Competency Porn and We Are Starving For It WeakC4 Flight Viz — Cockpit View A Mexican surveillance giant you’ve never heard of is now watching the U.S. border Surelock: Deadlock-Free Mutexes for Rust RISC-V 101 – what is it and what does it mean for Canonical? | Ubuntu The Problem That Built an Industry How Much Linear Memory Access Is Enough? | Solidean Investigating Split Locks on x86-64 Simplest hash functions Sybilproof reputation mechanisms (2005) [pdf] What is a property? How Complex is my Code? Static code analysis in Kotlin — tools overview Toffoli gates are all you need PGLite evangelism dcmake: a new CMake debugger UI Clojure on Fennel part one: Persistent Data Structures Fragments: April 2 Python Release Python install manager 26.1 The Life and Death of the Book Review - Liberties Introducing Database Traffic Control — PlanetScale Bitcoin miners are losing $19,000 on every BTC produced as difficulty drops 7.8% God sleeps in the minerals Building slogbox Apple Silicon and Virtual Machines: Beating the 2 VM Limit Who was “Not Even Wrong” first? Pokemon Evolution Vs Darwinian Evolution The APL Programming Language Source Code
Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7
Simon Willison · 2026-04-17 · via Hacker News

16th April 2026

For anyone who has been (inadvisably) taking my pelican riding a bicycle benchmark seriously as a robust way to test models, here are pelicans from this morning’s two big model releases—Qwen3.6-35B-A3B from Alibaba and Claude Opus 4.7 from Anthropic.

Here’s the Qwen 3.6 pelican, generated using this 20.9GB Qwen3.6-35B-A3B-UD-Q4_K_S.gguf quantized model by Unsloth, running on my MacBook Pro M5 via LM Studio (and the llm-lmstudio plugin)—transcript here:

The bicycle frame is the correct shape. There are clouds in the sky. The pelican has a dorky looking pouch. A caption on the ground reads Pelican on a Bicycle!

And here’s one I got from Anthropic’s brand new Claude Opus 4.7 (transcript):

The bicycle frame is entirely the wrong shape. No clouds, a yellow sun. The pelican is looking behind itself, and has a less pronounced pouch than I would like.

I’m giving this one to Qwen 3.6. Opus managed to mess up the bicycle frame!

I tried Opus a second time passing thinking_level: max. It didn’t do much better (transcript):

The bicycle frame is entirely the wrong shape but in a different way. Lines are more bold. Pelican looks a bit more like a pelican.

I don’t think Qwen are cheating

A lot of people are convinced that the labs train for my stupid benchmark. I don’t think they do, but honestly this result did give me a little glint of suspicion. So I’m burning one of my secret backup tests—here’s what I got from Qwen3.6-35B-A3B and Opus 4.7 for “Generate an SVG of a flamingo riding a unicycle”:

I’m giving this one to Qwen too, partly for the excellent <!-- Sunglasses on flamingo! --> SVG comment.

What can we learn from this?

The pelican benchmark has always been meant as a joke—it’s mainly a statement on how obtuse and absurd the task of comparing these models is.

The weird thing about that joke is that, for the most part, there has been a direct correlation between the quality of the pelicans produced and the general usefulness of the models. Those first pelicans from October 2024 were junk. The more recent entries have generally been much, much better—to the point that Gemini 3.1 Pro produces illustrations you could actually use somewhere, provided you had a pressing need to illustrate a pelican riding a bicycle.

Today, even that loose connection to utility has been broken. I have enormous respect for Qwen, but I very much doubt that a 21GB quantized version of their latest model is more powerful or useful than Anthropic’s latest proprietary release.

If the thing you need is an SVG illustration of a pelican riding a bicycle though, right now Qwen3.6-35B-A3B running on a laptop is a better bet than Opus 4.7!