πŸ“° AI Dev Weekly

The week's biggest AI developer news, with my take

Every Thursday
  1. #14

    AI Dev Weekly #14: Claude Fable 5 Controversy, DiffusionGemma Breaks Text Generation, Apple Rebuilds Siri

    Week of June 5-11, 2026: Anthropic ships Fable 5 with hidden competitor blocking, Google invents text diffusion at 1000 tok/s, Apple announces Core AI + Xcode 27, and a German court makes AI providers liable.
  2. #13

    AI Dev Weekly #13: Microsoft Declares Independence β€” 7 In-House Models, Kills Claude Code, RTX Spark Dev Box

    This week: Microsoft Build 2026 drops 7 homegrown AI models (no OpenAI data), ends Claude Code licenses, launches Surface RTX Spark Dev Box. Plus MiniMax M3, NVIDIA Computex, and Grok's next model teased.
  3. #12

    AI Dev Weekly #12: Opus 4.8 Drops, Anthropic Hits $965B, Chinese AI Goes 99% Cheaper, Microsoft Builds Its Own Coding Model

    This week: Claude Opus 4.8 tops every benchmark, Anthropic surpasses OpenAI in valuation, DeepSeek and MiMo slash prices by 99%, Microsoft announces its own coding model for Build 2026, and StepFun drops a 400 t/s open-weight model.
  4. #11

    AI Dev Weekly #11: Google I/O Drops Gemini 3.5 Flash, Kills Gemini CLI, and Karpathy Joins Anthropic

    This week: Google I/O 2026 reshapes the AI coding landscape with Gemini 3.5 Flash and Antigravity 2.0, Gemini CLI gets a June 18 death date, quota backlash leads to a 3x permanent boost, Karpathy defects to Anthropic, and a massive npm supply chain attack hits 323 packages.
  5. #10

    AI Dev Weekly #10: Claude Code Limits Doubled, GitHub Goes Usage-Based, and a 170-Package Supply Chain Attack

    This week: Anthropic doubles Claude Code rate limits after SpaceX compute deal, GitHub Copilot shifts to token-based billing June 1, and a coordinated attack compromises TanStack, Mistral AI SDK, and 170+ packages.
  6. #9

    AI Dev Weekly #9: Gemini 3.2 Flash Leaks Before I/O, GPT-5.5 Instant Becomes Default, and Enterprise Agents Go Self-Hosted

    This week: Google's unreleased Gemini 3.2 Flash outperforms 3.1 Pro on coding at $0.25/M tokens, OpenAI makes GPT-5.5 Instant the new ChatGPT default, and three companies launch self-hosted coding agents for enterprise.
  7. #8

    AI Dev Weekly #8: Mistral Medium 3.5 Goes Open-Weight, GPT-5.5 Lands in Codex, and Anthropic's $200 Billing Bug

    This week: Mistral drops a 128B open-weight flagship with cloud coding agents, GPT-5.5 replaces 5.4 at 40% less cost, and Anthropic charges a user $200 extra then refuses a refund.
  8. #7

    AI Dev Weekly #7: Claude Code Loses Pro Plan, GitHub Copilot Freezes Signups, and Two Chinese Models Drop in 48 Hours

    This week: Anthropic removes Claude Code from Pro, GitHub pauses all Copilot signups, Kimi K2.6 and MiMo V2.5 Pro launch within 48 hours, and flat-rate AI subscriptions are dying.
  9. Extra

    AI Dev Weekly Extra: Did Anthropic Let Opus 4.6 Rot So 4.7 Would Look Better?

    Opus 4.6 degraded for weeks. Now Opus 4.7 arrives with huge benchmark gains. Coincidence? Here's what actually happened and why it matters for developers.
  10. #6

    AI Dev Weekly #6: OpenAI's $852B Wobble, GPT-5.4 Solves 60-Year Math Problem, and Agents Get Infrastructure

    This week: OpenAI investors question the valuation while VCs throw $800B at Anthropic, GPT-5.4 Pro cracks an ErdΕ‘s conjecture, OpenAI ships agent sandboxes, Gemini CLI gets subagents, and a federal court rules your AI chats aren't privileged.
  11. #5

    AI Dev Weekly #5: Anthropic's Too-Dangerous Model, $30B Revenue, and China's GLM-5.1 Beats Everyone

    This week: Anthropic built Claude Mythos but won't release it, hit $30B revenue surpassing OpenAI, Meta shipped Muse Spark, and a Chinese open-source model topped the coding leaderboard.
  12. #4

    AI Dev Weekly #4: Anthropic Leaks Everything, OpenAI Raises $122B, and Qwen 3.6 Drops Free

    This week: Anthropic accidentally publishes Claude Code's entire source code to npm, OpenAI closes the largest private funding round in history, and Alibaba drops Qwen 3.6 Plus for free on OpenRouter.
  13. #3

    AI Dev Weekly #3: Claude Code Goes Auto, Cursor's Chinese Secret, and GitHub Wants Your Data

    This week: Anthropic ships auto mode and Discord/Telegram channels for Claude Code, Cursor gets caught building Composer 2 on a Chinese open-source model, and GitHub will train on your Copilot data by default.
  14. Extra

    AI Dev Weekly Extra: Xiaomi's Trillion-Parameter 'Hunter Alpha' Was Never DeepSeek V4

    A mystery AI model appeared on OpenRouter with no attribution. Everyone assumed it was DeepSeek V4. It was Xiaomi. Here's why that matters for developers.
  15. #2

    AI Dev Weekly #2: Garry Tan's 'God Mode', Cursor Composer 1.5, and Anthropic Finds Firefox Bugs

    This week: Y Combinator's CEO shares his Claude Code setup and the internet loses its mind, Cursor ships RL-trained self-summarization, and Claude Opus finds 22 Firefox vulnerabilities.
  16. #1

    AI Dev Weekly #1: Claude Code Takes the Crown, Musk Raids Cursor

    This week's biggest AI developer news: Claude Code is now the most-used coding tool, xAI poaches two senior Cursor engineers, and 95% of devs now use AI weekly.