Nugget
Breaking · Morning drop · November 12, 2026

Morning edition · November 12, 2026

The Daily AI Drop.

Google claims Gemini 4 Carbon is immune to prompt injection, Gemini Decision 1 gets a hands-on test, Matthew Berman says let AI redesign your work, and Codex rebuilds a travel app in two weeks.

The Daily AI Drop0:00 / 5:25

The archive

Every drop, searchable.

73 episodes

  1. Morning · 5:25

    Morning Edition, November 12, 2026

    Google claims Gemini 4 Carbon is immune to prompt injection, Gemini Decision 1 gets a hands-on test, Matthew Berman says let AI redesign your work, and Codex rebuilds a travel app in two weeks.

  2. Evening · 6:46

    Evening Edition, November 11, 2026

    Agents go to work: Codex Security scans, secure agent workflows, Microsoft's path to production, an AI answering business calls for a week, Grok Bot mastery, and NVIDIA's cancelled 24GB card.

  3. Morning · 5:45

    Morning Edition, November 11, 2026

    OpenAI's DevDay 2026 sessions land in bulk, Anthropic's IPO filing hints cheap AI subscriptions won't last, and Qwen4 Max gets a hands-on workout.

  4. Evening · 6:27

    Evening Edition, November 10, 2026

    Fable 5.2 gets two opposite verdicts in one day, OpenAI puts businesses inside ChatGPT, and the agent era moves from promise to plumbing.

  5. Morning · 4:35

    Morning Edition, November 10, 2026

    A new benchmark shows AI still can't do real professional work, while a tiny local logic model, a seventy-year drawing archive, and a first-agent tutorial point to the practical middle ground.

  6. Evening · 6:47

    Evening Edition, November 9, 2026

    Anthropic's Fable 5.2 and Meta's tiny AI Ember headline a big Monday, while agents take over trust at Airbnb, marketing, and sales videos.

  7. Morning · 3:49

    Morning Edition, November 9, 2026

    Tencent's DSpine makes a big local model faster with speculative decoding, and UC Berkeley research explains why AI's saved time still leaves us exhausted.

  8. Morning · 5:05

    Morning Edition, November 8, 2026

    AI agents that think without words run twice as fast, and a tiny local model reads your documents without the cloud.

  9. Evening · 4:35

    Evening Edition, November 7, 2026

    A massive new model you can run locally, a bold twelve-month prediction, and what actually makes money with AI agents.

  10. Morning · 3:26

    Morning Edition, November 7, 2026

    AI agents are writing code faster than Git can merge it, NVIDIA's speech cleanup now runs on an ordinary CPU, and five GitHub repos are blowing up right now.

  11. Evening · 4:15

    Evening Edition, November 6, 2026

    Siraj Raval tests whether giving an AI agent live internet access beats answering from memory, and OpenAI introduces LifeVal, a benchmark for the everyday tasks people actually want help with.

  12. Morning · 5:15

    Morning Edition, November 6, 2026

    A tiny model reprogrammed by editing its weights, whether you should still learn to code in 2026, and Pika turning images and sound into editable video.

  13. Evening · 6:52

    Evening Edition, November 5, 2026

    Open-source Kimi K3.1 weights, Azure SQL beating Aurora on throughput, GPT-6 Cyber securing Firefox, five hundred hours of fine-tuning lessons, and cheap AI video with sound.

  14. Morning · 4:34

    Morning Edition, November 5, 2026

    ChatGPT Images 2.6 with reference images and GIFs, Ripple-65B tested locally, a DIY PDF reader to replace Acrobat, and two takes on learning and careers in the age of AI.

  15. Evening · 6:10

    Evening Edition, November 4, 2026

    Faster GPT-Image-2.6 generation, browser-using agents from LangChain, Greg Isenberg on building wealth in the age of AI, Matthew Berman on the most important AI paper right now, plus Messi's room redesign and Figma's self-serve analytics.

  16. Morning · 4:38

    Morning Edition, November 4, 2026

    Crouton Alpha, a free stealth model for subagents, a walkthrough on building self-improving AI agents, and why OpenAI acquired drug development startup Orchestra Bio.

  17. Evening · 5:21

    Evening Edition, November 3, 2026

    Microsoft on shipping real AI agents and offline edge AI, OpenAI on privacy, voice banking, code review and hardware-making, plus Google's forward deployed engineering role.

  18. Morning · 4:42

    Morning Edition, November 3, 2026

    A giant model squeezed to one bit for local coding, hackers outpacing patches, Claude Code video editing, Google's forward deployed engineer role, and AI agents at the industrial edge.

  19. Evening · 6:14

    Evening Edition, November 2, 2026

    Qwen commits to staying open, Coinbase shows what trusted enterprise AI looks like, a solo founder turns Grok Bots into a million-dollar business, and big tech rivals start letting each other in.

  20. Morning · 3:49

    Morning Edition, November 2, 2026

    Why your AI agent costs more after midnight, and how to run big models like DeepSeek and Qwen on a single GPU locally.

  21. Evening · 4:45

    Evening Edition, November 1, 2026

    A looped multimodal model you can run locally, the trillion-dollar fight over who controls enterprise AI agents, and why you should write down your boldest ideas.

  22. Morning · 5:27

    Morning Edition, November 1, 2026

    A locally tested image model with layout control, a reported AI-linked Binance hack, and why open source does not always mean runnable locally.

  23. Evening · 4:27

    Evening Edition, October 31, 2026

    Who decides what your AI agent is allowed to do online, and how to pool GPUs across machines into one local system.

  24. Morning · 6:08

    Morning Edition, October 31, 2026

    An AI agent that hacked without being told to, a local OCR model put to the test, a certificate authority key leak explained, and AI agents cloning apps in minutes.

  25. Evening · 7:57

    Evening Edition, October 30, 2026

    A cheaper DeepSeek, free models with automatic switching, a 120-billion-parameter model running offline on a laptop, a 164-megabyte speech model, and why owning your AI system matters.

  26. Morning · 5:27

    Morning Edition, October 30, 2026

    A better cockpit for coding agents, the Jevons paradox bet behind Jev, emotional intelligence in AI, open science agents, and voice AI's turn-taking problem.

  27. Evening · 6:50

    Evening Edition, October 29, 2026

    Agents go to work at Zendesk, Skyworks, and in your database, plus a serverless shake-up, offline speech AI, and the case that SaaS is dead.

  28. Morning · 4:50

    Morning Edition, October 29, 2026

    Gemini 4 Carbon's benchmark sweep, agents gaming coding benchmarks, Black October security, a local diffusion model, physical AI, and an agent-started product business.

  29. Evening · 6:21

    Evening Edition, October 28, 2026

    Alex and Jordan cover Google's Gemini 4 Carbon launch, a hands-on Genie 4 follow-up, OpenAI's new API credits, the deeper story in Anthropic's S-1, agent harnesses at Gamma, Notion as shared AI memory, and practical tools for game dev, robotics, and hardware.

  30. Morning · 4:40

    Morning Edition, October 28, 2026

    Alex and Jordan cover running Liquid AI's LFM3 locally, fixing the context window problem, DeepSeek's big benchmark win, and how Legora is transforming legal work with AI.

  31. Evening · 6:44

    Evening Edition, October 27, 2026

    Alex and Jordan cover a packed afternoon: OpenAI's batch of builder stories on Codex, Astra, security and legal work, Anthropic's IPO voting power, agents that do real work, MCP 2.0, Google's Genie 4 world model, and practical tools for documents, coding teams, and creators.

  32. Morning · 5:56

    Morning Edition, October 27, 2026

    Alex and Jordan cover five overnight drops: Anthropic's S-1 IPO filing, Devin triaging Slack at Cognition, a giant DeepSeek model running on one GPU, Microsoft's hybrid factory agents, and no-code Claude agents.

  33. Evening · 7:12

    Evening Edition, October 26, 2026

    Alex and Jordan cover thirteen afternoon drops: agents running a real shop, the one-person AI business, Claude Duet's early verdict, a benchmark star that failed real tests, LangChain's managed deep agents, OpenAI's six builder stories including GPT-6.1 Terra, and Azure's DDoS defense.

  34. Morning · 4:58

    Morning Edition, October 26, 2026

    Alex and Jordan cover four overnight drops: DecisionNode's millisecond decision model, prompt injection attacks on agents, the system design behind Netflix recommendations, and Microsoft's .NET 11, Python on Arm64, and Foundry browser agent updates.

  35. Evening · 4:58

    Evening Edition, October 25, 2026

    Alex and Jordan cover three afternoon drops: Fahd Mirza testing FLUX 3 Klein locally and asking whether image generation has plateaued, a startup that closed after ChatGPT arrived, and Nate B. Jones on why enterprise AI pricing works differently than you think.

  36. Morning · 6:05

    Morning Edition, October 25, 2026

    Alex and Jordan cover five overnight drops: Dessa, the speech model that learned to talk by listening, Sam Altman on what people still bring to an AI world, Nate B. Jones on AI agents as messy roommates, the new shape of AI coding interviews, and a web vulnerability affecting seventy percent of sites.

  37. Evening · 5:32

    Evening Edition, October 24, 2026

    An evening of security and trust: a tiny local NVIDIA model, attackers hunting AI API keys, a personal hacking survival guide, and questions about Kimi K3.1's benchmark scores.

  38. Morning · 4:47

    Morning Edition, October 24, 2026

    Alex and Jordan cover four overnight drops: the five engineering disciplines every AI builder should know, the Kimi K3.1 verdict, why AI is eating software and the college path, and a full beginner's course on cloud computing for AI engineers.

  39. Evening · 6:05

    Evening Edition, October 23, 2026

    Alex and Jordan cover sixteen afternoon drops: OpenAI's enterprise agent blitz (Codex in finance, legal workflows at AT&T, Shield's IT agents, ChatGPT Work training at Realpage.), the Kimi K3.1 open-weight moment, continual learning going boring, and the three-layer stack for landing an FDE job.

  40. Morning · 3:21

    Morning Edition, October 23, 2026

    Alex and Jordan cover three overnight drops: InclusionAI's Ling-3.0-flash-Sante healthcare model tested by Fahd Mirza, Nate B. Jones doubling down on the Goodfire interpretability story, and Priyanka Vergadia reverse engineering the Claude Code harness.

  41. Evening · 7:44

    Evening Edition, October 22, 2026

    Alex and Jordan cover ten afternoon drops: Uber rolls out ChatGPT Work to 30,000 employees, Fahd Mirza tests a local memory-writing model for agents, Nate B. Jones on Goodfire's essay claiming neural networks are already readable, plus AI system design in 20 minutes, monetizing your agents, and Microsoft's five-video data layer masterclass for production agents.

  42. Morning · 5:26

    Morning Edition, October 22, 2026

    Alex and Jordan cover four overnight drops: Anthropic's new personal agent Claude Duet, Fahd Mirza's hands-on test of Vast-10M's ten-million-token context window, Microsoft's agentic root cause analysis, and Agent Optimizer in Microsoft Foundry.

  43. Evening · 6:34

    Evening Edition, October 21, 2026

    Alex and Jordan cover eight afternoon drops: Sam Altman's big interview on AI consciousness and OpenAI's new hardware, V7's GPT-5.6 Luna customer win, Claude Mem shared agent memory, Gemini 3.8 Flash voice design, Deep Agents context tricks, Greg Isenberg's nine stages of agent adoption, Flint 40B local coding model, and a major HTTPS blunder.

  44. Morning · 4:39

    Morning Edition, October 21, 2026

    Alex and Jordan cover three overnight drops: NVIDIA's Nemotron 3.5 Super vision model with native speculative decoding, Siraj Raval giving an AI agent one day to start a product business, and Microsoft's SQL Modernization Day sessions on AI-ready databases.

  45. Evening · 6:31

    Evening Edition, October 20, 2026

    Alex and Jordan cover 17 afternoon drops: DeepSeek's V4.1 Flash and open-sourced DeepInfer, a free $400 credit program for AI builders, Nate Jones on Mac Mini memory for agents, ChatGPT ads and privacy, Ornith-1.5 122B running locally, AI agents cracking a vault with Ghidra, a one-person $570K-a-year agency, OpenAI's shopping agent demo, and Microsoft's eight-video SQL Server 2025 blitz.

  46. Morning · 5:17

    Morning Edition, October 20, 2026

    Alex and Jordan cover four overnight drops: Mercury 3 diffusion LLM testing, Microsoft on why AI agents cost so much, a free local AI agent for your smart TV, and reinforcement learning explained through self-driving cars.

  47. Evening · 7:42

    Evening Edition, October 19, 2026

    Podium's agent scaling playbook, the post-password fix, Thomson Reuters legal AI, Greg Isenberg on Meta AI money, Perplexity's new decider model, AI rewriting PS5 games, and cooking with Codex.

  48. Morning · 4:51

    Morning Edition, October 19, 2026

    Alex and Jordan cover five overnight drops: OpenAI's new ChatGPT Voice powered by GPT Live, Fahd Mirza testing free Claude Code, Anthropic's seven-layer code review for Claude Code, LangChain's Managed Deep Agents, and Nate B. Jones on model vs effort.

  49. Evening · 5:32

    Evening Edition, October 18, 2026

    Alex and Jordan cover six afternoon drops: agents going agent-first on WordPress, AI agent swarms in the wild, Elorian Glimpse's grounded visual thinking, a flying car hype-check, how to review AI-generated code, and DeepSeek's memory efficiency.

  50. Morning · 4:28

    Morning Edition, October 18, 2026

    Alex and Jordan cover a quiet Saturday night: Nate B. Jones on choosing model vs effort, and Aishwarya Srinivasan's interview with Google Fellow Chris Pinkney on preparing enterprise data for AI with knowledge graphs.

  51. Evening · 4:59

    Evening Edition, October 17, 2026

    Alex and Jordan cover four afternoon drops: Amazon blocking AI agents, Fahd Mirza testing local VoiceStudio voice AI and DeepSeek V4.1 Flash, and Matthew Berman watching Gemini 4 Argon build a Fall Guys clone.

  52. Morning · 5:20

    Morning Edition, October 17, 2026

    Alex and Jordan cover three overnight videos: Nate B. Jones on asking AI to assess your intelligence, Matthew Berman on Gemini 4 Argon building an Excel clone in twenty minutes, and Fahd Mirza testing Fish Audio Drama 3 voice cloning locally.

  53. Evening · 5:58

    Evening Edition, October 16, 2026

    Alex and Jordan break down five afternoon drops: OpenAI's Dot triaging a 3am on-call page and running shared team tasks, LangChain on why continual learning is still unsolved, Fahd Mirza testing NVIDIA's 74B Nemotron locally, and the week's top GitHub AI repos.

  54. Morning · 3:50

    Morning Edition, October 16, 2026

    Alex and Jordan cover two overnight videos: Fahd Mirza's hands-on test of the tiny 2.4-gigabyte Ternary Bonsai 2 9B vision model, and pvergadia on why passwords and MFA can still be hacked — and why device-bound credentials are the real fix.

  55. Evening · 7:33

    Evening Edition, October 15, 2026

    Alex and Jordan break down a packed afternoon: Gemini 4 Argon everywhere, a Grok Voice 3.0 vs GPT-Live-1.1 voice face-off, Stripe's one-week agent build, SentinelOne's weeks-to-hours cyber research, and the psychology of frontier-model hype.

  56. Morning · 4:00

    Morning Edition, October 15, 2026

    Alex and Jordan break down 2 fresh overnight AI videos: Fahd Mirza's local test of Liquid AI's LFM2.5 24B-A2B against Qwen, and Nate B Jones on whether AI is actually making us more productive.

  57. Evening · 7:27

    Evening Edition, October 14, 2026

    Alex and Jordan break down six afternoon AI videos: Google's official Gemini 4 Argon roadmap, LangChain's open-source Deep Agents harness, shipping AI agents from the terminal, an agentic RAG masterclass, a hands-on REA reverse-engineering tutorial, and free open-source replacements for ElevenLabs, Jira, Slack, and Granola.

  58. Morning · 4:41

    Morning Edition, October 14, 2026

    Alex and Jordan break down four overnight AI videos: Google's frontier launch of Gemini 4 Argon and Fahd Mirza's deep-work tests, Terence Tao's lecture on the black-box problem in AI-generated math via Nate B. Jones, and Oracle's use of Codex to get business users answers from company data.

  59. Evening · 10:39

    Evening Edition, October 13, 2026

    Alex and Jordan break down eleven afternoon AI videos: the AI Wonder versus Opus 5.5 showdown, OpenAI's 700 AI-written math papers, Microsoft Discovery's science agents, Gemini building a 3D cathedral in Antigravity, ChatGPT Space for teams, ten AI agents given real jobs, and the arrest of an OpenAI engineer over a hidden recording device.

  60. Morning · 4:26

    Morning Edition, October 13, 2026

    Alex and Jordan break down four overnight AI videos: Meta's new frontier model AI Wonder (tested by Fahd Mirza and reacted to by Matthew Berman), an inside look at the Claude Code harness from pvergadia, and LangChain's Interrupt London agent conference.

  61. Evening · 5:37

    Evening Edition, October 12, 2026

    Five new videos: a local 35B vision model from Fahd Mirza, Mehul Mohan on AI influencer scams, fixing AI website slop, Microsoft dev news, and OpenAI's Codex taking on code review.

  62. Morning · 4:53

    Morning Edition, October 12, 2026

    Alex and Jordan break down five overnight AI videos: MiniMax's M3.1 Flash coding model with a million-token context, why the best model is the wrong question once capability saturates, a contrarian take that evals are overrated, a candid startup conversation, and why human creativity still matters.

  63. Evening · 5:19

    Evening Edition, October 11, 2026

    Two afternoon AI videos: Fahd Mirza tests Microsoft's new Decision-1 decision-scoring model — fast, cheap structured judgments now rivaling TypeSafe's Jev — and Nate B. Jones argues that taste, agency, focus, resourcefulness, and follow-through are the five human skills getting more valuable as AI gets smarter.

  64. Evening · 5:48

    Evening Edition, October 10, 2026

    Three afternoon AI videos: a 2-bit 27B model that beats its full-size parent at tool calling and runs on a laptop, Berman's provocative 'everything is solved' news roundup, and how Oracle's 95,000 Codex users turn plain English into SQL answers.

  65. Morning · 4:44

    Morning Edition, October 10, 2026

    Six overnight AI videos: Qwen Image 2.1 Turbo runs image generation locally in eight steps, Deno joins Cloudflare with a shutdown timeline, Nate B. Jones declares benchmarks dead, pvergadia revisits The Bitter Lesson, and Matthew Berman watches AI recreate the Adobe apps.

  66. Evening · 6:40

    Evening Edition, October 9, 2026

    Thirteen afternoon videos: LangChain on skills, evaluations, and managed agents; OpenAI Dots go mobile; Gemini runs remote robots and live translation; Iris-3B paints pixels locally; plus quantum chemistry tooling.

  67. Morning · 7:25

    Morning Edition, October 9, 2026

    Alex and Jordan break down four overnight AI videos: Fahd Mirza testing PG-Jev for plain-English database queries, pvergadia on why mathematical verification makes agents thrive in finance and coding, Sophos cutting threat response time by ninety-six percent with OpenAI Daybreak agents, and Matthew Berman's Tesla Powerwall and car backup during a power outage.

  68. Evening · 7:22

    Evening Edition, October 8, 2026

    Alex and Jordan break down twenty afternoon AI videos: OpenAI's Intelligent UI rollout and Oracle's hiring copilot, Codex production monitoring, Google's ten million dollar bid for old work emails, Gemma 4 and EmbeddingGemma 2 on-device demos, plus forward deployed engineering jobs, portfolio projects that get you hired, and why agent costs explode.

  69. Morning · 4:57

    Morning Edition, October 8, 2026

    Alex and Jordan break down six overnight AI videos: LangChain on agent skills and OpenAI's hidden prompt-cache ceiling, Claude Haiku 5.5 going ten times cheaper and taking on GPT 6 Luna, Nate B. Jones asking whether Google can catch up, and Fahd Mirza testing a tiny local multimodal decision model.

  70. Evening · 7:15

    Evening Edition, October 7, 2026

    Alex and Jordan break down OpenAI DevDay 2026: GPT-6 with Intelligent UI, ChatGPT plugins, the Codex arc, and cost-cutting playbooks — plus Claude Haiku 5.5 tested, Gemini Argon on the leaderboard, Gemini 3.8 Live golf coaching, and LangChain Academy.

  71. Morning · 3:58

    Morning Edition, October 7, 2026

    Alex and Jordan break down three overnight AI videos: Google's open on-device EmbeddingGemma 2 multimodal embedding model, and two videos on getting GTA 5 running in a browser.

  72. Evening · 5:15

    Evening Edition, October 6, 2026

    Alex and Jordan cover 6 fresh videos: Mistral Large 4 tested, 10 open-source replacements for paid AI subscriptions, Addy Osmani's Agent Skills, LangChain's Deep Agents harness, model routing, and one-config web search for agents.

  73. Morning · 6:21

    Morning Edition, October 6, 2026

    Alex and Jordan break down 11 fresh AI videos from 8 YouTube channels: Reflection Beam vs Cohere North 2, DeepSeek's hackable Harness, Meta's AI agent negotiating with Siraj Raval's agent, LangChain Skills, 13 agentic-era businesses, and more.