Links / November 2026

Sunday, 8 November 2026

The AI links the most newsletters and link sources had shared by the end of the day (UTC), each source counted once.

As of 2026-11-08 23:59 UTC Status sealed

That day 36-hour window

  1. Shared by 1

    Build a coding agent

    “— Mihail builds a coding agent from scratch.” via Latent Space / AINews

    Reported byLatent Space / AINews

    mihaileric.com First shared 2026-11-08
  2. Shared by 1

    research/sqlite-parse-number-c-vs-rust at main

    A fast C path beats Rust by 2x speed while using 20x less code size for SQLite number parsing. Includes speed tables showing 67 ns for plain C versus 19 ns for the optimized C version. cited

    Reported bySimon Willison

    github.com analysis · 5 min read First shared 2026-11-08
  3. Shared by 1

    Claude

    “…ummed and sorted. I fired up regular Claude chat (Opus 5.5) and prompted: I'm thinking about a SQLite custom function 'parse_number(value)' that strips ',', '$', '£' and '%', handles a leading '-', and returns 'NULL' for anything it can’t parse Explore the difference between building this as a SQLite extension in C v.s. in Rust - I want to compare…” via Simon Willison

    Reported bySimon Willison

    claude.ai 1 min read First shared 2026-11-08
  4. Shared by 1

    rusqlite/rusqlite: Ergonomic bindings to SQLite for Rust

    Rusqlite version 0.41.0 provides ergonomic SQLite bindings for Rust with bundled support for SQLite 3.54.0. Includes Cargo features for SQLCipher, window functions, and virtual tables, plus instructions for configuring bundled builds. cited

    Reported bySimon Willison

    github.com repo · 8 min read First shared 2026-11-08
  5. Shared by 1

    asg017/sqlite-loadable-rs: A framework for writing fast and performant SQLite extensions in Rust

    sqlite-loadable-rs is a beta Rust framework for building fast, loadable SQLite extensions. Includes benchmarks showing Rust extensions are 10-15% slower than C but 20-30x faster than Go. cited

    Reported bySimon Willison

    github.com repo · 6 min read First shared 2026-11-08
  6. Shared by 1

    simonw/research: Research projects

    This GitHub repo hosts research projects where every line of text and code was written by an LLM. Includes specific projects like a Rust-based SQLite extension for SPARQL 1.1 and a webframe scraper evaluation. cited

    Reported bySimon Willison

    github.com repo · 76 min read First shared 2026-11-08
  7. Shared by 1

    AIE Code Summit 2026, Tue, Nov 10th, 2026 @ 8:00 AM ai.engineer

    “…scriber and join us! $400 code for LS paid subscribers: Use LS30 at checkout for 30% off :) AI News for 10/28/2026-10/29/2026. We checked 12 subreddits, 544 Twitters and no further Discords. AINews' website lets you search all past issues. As a reminder, AINews is now a section of Latent Space. You can opt in/out of email frequencies!AI Twitter RecapMistral…” via Latent Space / AINews

    Reported byLatent Space / AINews

    app.ai.engineer 1 min read First shared 2026-10-30
  8. Shared by 1

    Understanding is the new bottleneck

    “— Geoffrey Litt’s approach to learning what changed in an AI-written diff.” via Latent Space / AINews

    Reported byLatent Space / AINews

    geoffreylitt.com 9 min read First shared 2026-11-08
  9. Shared by 1

    The AI Productivity Paradox Research Report

    “How to Kill the Code Review and the underlying Faros study — Teams with high AI adoption merged 98% more PRs, while their PR review time was 91% longer.” via Latent Space / AINews

    Reported byLatent Space / AINews

    faros.ai 5 min read First shared 2026-11-08
  10. Shared by 1

    How We Made Our Monorepo Ergonomic for Agents

    Basis rewrote 20 AGENTS.md files and fixed 9 projects to boost commit velocity by 2.5x. Details the six-layer architecture separating canonical truth from non-canonical history to prevent agent confusion. cited

    Reported byLatent Space / AINews

    getbasis.ai news report · 12 min read First shared 2026-11-08
  11. Shared by 1
    mihaileric.com First shared 2026-11-08
  12. Shared by 1

    Build or Buy? Agent-Developer Workspaces

    Monaco built an in-house agent workspace system using Coder and internal AMIs to run parallel agents securely. Details the architecture with one app per VM, a golden database clone strategy, and specific costs like EKS nodes and NAT charges. cited

    Reported byLatent Space / AINews

    monaco.com news report · 6 min read First shared 2026-11-08
  13. Shared by 1

    Minions: Stripe’s one-shot, end-to-end coding agents

    Stripe's unattended coding agents merge over a thousand pull requests weekly with no human-written code. Details the devbox environment, Block's goose fork, and the two-round CI limit for fixing flaky tests. cited

    Reported byLatent Space / AINews

    stripe.dev announcement · 7 min read First shared 2026-11-08
  14. Shared by 1
    x.com X post First shared 2026-11-08
  15. Shared by 1

    Mihail on The Thinking Mode

    “… When code review becomes the bottleneck1:20:37 Security, dependencies and permissions1:30:13 Background agents and Stripe’s Minions1:34:16 MCP portals and why infrastructure converges1:41:22 Building the software factory1:45:24 What work is left for developers?1:52:16 From personal adoption to team autonomy Opening excerpt from The Thinking Mode, EO…” via Latent Space / AINews

    Reported byLatent Space / AINews

    youtube.com Video First shared 2026-11-08
  16. Shared by 1
    nytimes.com First shared 2026-11-08
  17. Shared by 1

    Boris Cherny’s Steps of AI Adoption

    “— The five stages we pull up on screen near the end of the episode.” via Latent Space / AINews

    Reported byLatent Space / AINews

    x.com X post First shared 2026-11-08
  18. Shared by 1

    Artificial Analysis Decision Index v2

    Artificial Analysis launches Decision Index v2, weighting five evaluations including AA-Paperclip at 30% and EvalSafe-O*NET-AA at 15%. Includes tables showing Decision Score metrics, Brier Score calibration data, and cost per 1M Tasks for the new index. cited

    Reported byAlphaSignal

    artificialanalysis.ai benchmark · 6 min read First shared 2026-10-13
  19. Shared by 1

    AA-Paperclip Benchmark

    “30% Decisions and extraction across 1,500 private resumes, receipts, invoices, and machine-learning papers. Each document includes five decisions and 10 facts, with all 15 answers required for a pass.” via AlphaSignal

    Reported byAlphaSignal

    artificialanalysis.ai 5 min read First shared 2026-11-08
  20. Shared by 1

    AA-Symmetry Benchmark

    Artificial Analysis launches AA-Symmetry, a composite benchmark aggregating nine evaluations to measure AI capabilities in knowledge work, coding, and science. Includes details on the Harvey Legal Agent Benchmark, which checks every deliverable against source documents so work with material hallucinations receives no credit. cited

    Reported byAlphaSignal

    artificialanalysis.ai benchmark · 5 min read First shared 2026-11-08

The week to that day two or more sources

  1. Shared by 7

    Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips

    Google makes Gemini Omni 1.2 Flash generally available at $0.07 per second for 720p video. Includes native 15-second generation, multi-turn conversational editing, and improved reference adherence for consistent characters and products. cited

    Reported byAlphaSignal, Latent Space / AINews, The Rundown AI, TLDR AI, Superhuman AI, The Neuron

    Discussed onHacker News 367 pts · 253 comments

    Where it was shared
    blog.google announcement · 4 min read First shared 2026-11-05
  2. Shared by 7

    Introducing Personal Finance in Claude

    Anthropic launches Personal Finance in Claude for US Pro, Max, and Ultra subscribers to analyze up to two years of read-only transaction data. Includes examples of spending breakdowns, upcoming payment alerts, and APR calculations, while noting the tool cannot move or invest money. cited

    Reported byAlphaSignal, The Rundown AI, TLDR AI, Forward Future, Superhuman AI, The Deep View, The Neuron

    Where it was shared
    claude.com announcement · 3 min read First shared 2026-11-05
  3. Shared by 9

    ChatGPT Images 2.6

    “OpenAI released ChatGPT Images 2.6 with native GIF generation from text or photo prompts” via AlphaSignal

    Reported byAlphaSignal, The Neuron, Superhuman AI, Ben's Bites, The Rundown AI, FryAI, AI Breakfast, The Deep View, Mindstream

    How 1 source put it
    openai.com First shared 2026-11-04
  4. Shared by 5

    GPT-6 Cyber (13 minute read) OpenAI introduced GPT-6 Cyber, a cybersecurity model designed to identify and validate…

    “OpenAI just released GPT-6 Cyber, a version of GPT-6 Astra built to find security holes and write exploits: code that takes advantage of a specific bug. OpenAI rates it “Critical” for cybersecurity, the highest risk tier in its internal framework. Not exactly a subtle job description.” via The Neuron

    Reported byThe Neuron, The Rundown AI, TLDR AI, AI Breakfast, The Deep View

    How 1 source put it
    openai.com First shared 2026-11-06
  5. Shared by 4

    Independent Assessment of AI R&D Acceleration at Anthropic in October 2026

    “METR estimates Anthropic's model capabilities are now improving 2x faster than the March 2024 to October 2025 baseline.” via AlphaSignal

    Reported byAlphaSignal, The Rundown AI, TLDR AI, Latent Space / AINews

    How 1 source put it
    metr.org 53 min read First shared 2026-11-05
  6. Shared by 5

    OpenAI agreed to acquire Pierre

    “OpenAI is acquiring Pierre, the startup behind the code review tools inside Codex, and plans to add its Code Storage service to the platform as well.” via The Rundown AI

    Reported byThe Rundown AI, TLDR AI, AI Breakfast, The Neuron

    Discussed onHacker News 265 pts · 93 comments

    How 1 source put it
    openai.com First shared 2026-11-04
  7. Shared by 9

    Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights

    GPT-6 Astra edits 0.016% of Qwen3.5-0.8B weights to raise MATH-500 from 7.9% to 48.1% without training. Includes demos where Astra identifies games from raw weights and writes MinAtar agents beating PPO 100M steps. cited

    Reported byTuring Post, Latent Space / AINews, The Neuron, The Rundown AI, Superhuman AI, Simon Willison, AlphaSignal, AI Breakfast

    Discussed onHacker News 2,960 pts · 900 comments

    Where it was shared
    rikiyat.github.io paper · 18 min read First shared 2026-11-02
  8. Shared by 6

    Kimi K3.1 Pro

    “Moonshot AI released full open weights for Kimi K3.1 Pro on Hugging Face” via AlphaSignal

    Reported byAlphaSignal, Latent Space / AINews, The Neuron, TLDR AI, FryAI

    Discussed onHacker News 640 pts · 191 comments

    AboutKimi-K3.1-Pro

    How 1 source put it
    huggingface.co First shared 2026-11-04
  9. Shared by 5

    A thousand fewer paths to root”

    “OpenAI's autonomous pipeline found and exploited 948 unique Linux kernel privilege escalation bugs in one run.” via AlphaSignal

    Reported byAlphaSignal, TLDR AI, AI Breakfast, Latent Space / AINews

    Discussed onTechmeme

    How 1 source put it
    openai.com First shared 2026-11-04
  10. Shared by 4

    Ripple-65B Preview, a Frontier Open-Weight Model

    Ricursive releases Ripple-65B Preview, a model achieving 80.52 on TerminalBench 2.1 for $32M. Details the thin-and-deep architecture, synthetic data pipelines, and laddering experiments that enabled this efficiency. cited

    Reported byThe Rundown AI, TLDR AI, AlphaSignal

    Discussed onHacker News 341 pts · 127 comments

    Where it was shared
    ricursive.com announcement · 4 min read First shared 2026-11-04