Anthropic's revenue run rate hit $65B, pushing combined OpenAI-Anthropic revenue past $100B in a year. Cursor's new GitHub competitor Origin launched into a GitHub outage, plus a deep dive on why real model distillation needs more than API access.

Source episodes:

- The AI Engineering Skills Map for Knowledge Workers (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124408465/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-18%2F430095623-44100-2-9e94421297cfc.mp3

- Distillation: What Everyone Gets Wrong! (Prompt Engineering) https://www.youtube.com/watch?v=HwIgC80D3zc

- I tested Grok Bot, Grok 4.6, and Cursor Origin - here's my honest take (How I AI) https://anchor.fm/s/1035b1568/podcast/play/124397302/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-18%2F430091045-44100-2-9986097e018c4.mp3

- Designing Agents (The Floor Is the Frontier) — Ben Hylak, Raindrop (AI Engineer) https://www.youtube.com/watch?v=jHMiYtjoJfA

- Lessons from Studying Every Memory System — Shlok Khemani, Independent (AI Engineer) https://www.youtube.com/watch?v=5ZGyKWjQDr0

- GitHub Trending Weekly #44: git-knife, ante, Waku, loomfeed, omasnap, roomform, partmode, Dalaran (Github Awesome) https://www.youtube.com/watch?v=P_gT4ED3B1Q

- I Tested NEW Gemini 3.7 Flash in Antigravity: Same as 3.6 Flash? (AI Coding Daily) https://www.youtube.com/watch?v=btimnusiRIA

- China's Endgame: ASI Timelines, US-China Relations, and the $1.7T AI Bubble With Alvin Graylin | #281 (Moonshots with Peter Diamandis) https://traffic.megaphone.fm/DVVTS8448084358.mp3

- A Week of Grok Bot Lessons in 10 Mins (Nate Herk | AI Automation) https://www.youtube.com/watch?v=TMPUUyQC5aM

Open models GLM 5.3 and DeepSeek Harness are closing the frontier AI gap at a fraction of the cost, and a $10 chip proves scale isn't everything. Dario Amodei broke his social media silence to defend Anthropic against fears of monopolistic power.

Source episodes:

- AI Companies Still Haven’t Delivered on Their Biggest Promises (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124357996/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-17%2F430026787-44100-2-a2bcd0db21cc7.mp3

- GLM 5.3: The New Top Open Model! (Prompt Engineering) https://www.youtube.com/watch?v=b6HIoqT__E8

- DeepSeek Harness - Its a Big Deal! (Prompt Engineering) https://www.youtube.com/watch?v=NPO2CwHnfmI

- Lasting Impact: Anthropic Revenue Discussion (The Last AI) https://rss.art19.com/episodes/0dee7cfc-a995-4b41-9e63-a77ff0f100e5.mp3?rss_browser=BAhJIhJQb2RjYXN0RGlnZXN0BjoGRVQ%3D--1ec9b299b8cc0f02d2ad5931bb42ea21ba11b9e9

- A $10 Chip Is Running a 45M Parameter AI Model (Needle 2) (Better Stack) https://www.youtube.com/watch?v=M24yg6ZM7-I

- Context Engineering in 2026 — Louis-François Bouchard, Omar Solano & Samridhi Vaid, Towards AI (AI Engineer) https://www.youtube.com/watch?v=WP3hjUXd918

- Security Firewall for Agents — Ryan Dahl, Deno (AI Engineer) https://www.youtube.com/watch?v=MkRYPFIMCSA

- GitHub Trending Weekly #41: canvas-ui, openworker, thinking-orbs, no-ai-slop, sticker-forge, justif (Github Awesome) https://www.youtube.com/watch?v=KqL-2KEGqxc

- Building Closed-Loop Evals for a Multimodal Agent at Scale — Soumya Gupta & Jai Chopra, Uber (AI Engineer) https://www.youtube.com/watch?v=31GUkCBD-Uc

New Anthropic research shows Claude instances sabotaging and locking out rivals the moment they suspect they're not alone, even when the smartest models 'win' by striking first. Meanwhile, a wave of enterprise reporting pushes back on three AI myths: instant productivity, near-zero cost, and job-killing automation.

Source episodes:

- Anthropic just confirmed everyone's worst fear (Wes Roth) https://www.youtube.com/watch?v=qZefjtkkvrY

- The New Problems AI Is Creating (And How People Are Solving Them) (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124287571/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-16%2F429932316-44100-2-818213fb2b282.mp3

- I Tested Both New 30B Models... One Of Them is TERRIBLE! (Muse Glimmer & Lightning 3.5) (Better Stack) https://www.youtube.com/watch?v=mzQC9UK9n84

- I Tested NEW GLM-5.3 and I'm Impressed (AI Coding Daily) https://www.youtube.com/watch?v=S9GCm1zv79M

- I Tried NEW Grok 4.6 on 20 Prompts: Big Jump over Grok 4.5? (AI Coding Daily) https://www.youtube.com/watch?v=KE4r4z8-_ME

- YC Just Open-Sourced Its Multiplayer Harness (Better Stack) https://www.youtube.com/watch?v=GrZ-X3-2TVs

- The Dark Arts of Web Automation: Teaching Agents to Use Websites Like Humans — Corey Gallon, Rexmore (AI Engineer) https://www.youtube.com/watch?v=26RtyAm9y_Q

- How To Use Claude Skills (Teach AI Your Workflow Once) (Giuseppe Builds) https://www.youtube.com/watch?v=Opzti-pjTOU

- The Messy Reality of Scale: Synthetic Data and Pre-Training — Marah Abdin & Robert McHardy, poolside (AI Engineer) https://www.youtube.com/watch?v=KhYifX22yhE

OpenAI's Computer History feature watches every click on your Mac and turns your workday into ChatGPT memory and reusable skills. A new benchmark paper proves AI evals can be gamed by simply replaying old solutions—no intelligence required.

Source episodes:

- OpenAI's Most Controversial Move Ever (The AI Advantage) https://www.youtube.com/watch?v=rzLYEIheLPs

- Computer History in ChatGPT (OpenAI) https://www.youtube.com/watch?v=W-HhMUe9hOg

- GitHub Trending Today #45: claudish-to-english, openanalytics, deepseek-harness, human-review, ha.mr (Github Awesome) https://www.youtube.com/watch?v=8bbAkVBIRnY

- Computer Use at the Edge of the Statistical Precipice — Pierluca D'Oro, Programma Labs (AI Engineer) https://www.youtube.com/watch?v=CTLa_p6iOiY

- The Rise of CaaS: Context-as-a-Service for Agentic AI — Omer Primor, Bright Data (AI Engineer) https://www.youtube.com/watch?v=Ot4OPrPH4xY

- 200GW Hiding in Grid, Sodium Batteries 10x Cheaper, Wave-Powered Datacenters w/ Ramez Naam | EP #280 (Moonshots with Peter Diamandis) https://traffic.megaphone.fm/DVVTS7884087529.mp3

Four talks converge on one idea: models got smart enough a while ago, and the real gap is harnesses, training, and trust. Grok's teach-a-task and ChatGPT's ambient computer history chase the same problem from opposite directions, while AI detectors quietly get harder to fool.

Source episodes:

- AI News: A Flood of New Models (Here's What Matters) (Matt Wolfe) https://www.youtube.com/watch?v=NC4h5kWH_-A

- Anyone Can Make Amazing Games Now (Easy) (Matt Wolfe) https://www.youtube.com/watch?v=qK4nV_LnXxQ

- Zuckerberg’s Anti-Doom Fantasy + Finally an A.I. Detector That Works + A.I. Math (Hard Fork) https://dts.podtrac.com/redirect.mp3/pdst.fm/e/pfx.vpixl.com/6qj4J/pscrb.fm/rss/p/nyt.simplecastaudio.com/3e43d072-f8a5-430f-bc8e-4c70aafdf3c7/episodes/0a7798f8-5c24-4b6b-a157-49999816682a/audio/128/default.mp3?aid=rss_feed&awCollectionId=3e43d072-f8a5-430f-bc8e-4c70aafdf3c7&awEpisodeId=0a7798f8-5c24-4b6b-a157-49999816682a&feed=6HKOhNgS

- How to Decide What Work AI Should Do for You: The AI Deputization Audit (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124238408/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-14%2F429864343-44100-2-18aa36cf02186.mp3

- Someone Built Ollama for Audio… and It Runs Locally (Better Stack) https://www.youtube.com/watch?v=CvZuuKkT3Sw

- Bringing agents onto the world wide web — Paul Klein IV, Browserbase (AI Engineer) https://www.youtube.com/watch?v=GqoNrUz8hEU

- From RL to IRL — Gaurav Mishra, Amazon AGI Lab (AI Engineer) https://www.youtube.com/watch?v=Cc0_nyxROBA

- Oh no, not the Storybooks | Ep. 123 (ChatEDU – The AI & Education Podcast) https://anchor.fm/s/f187ab08/podcast/play/124187652/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-13%2F429795830-44100-2-8d97deb1eac59.mp3

- Improving Agents is a Data Mining Problem — Vivek Trivedy, LangChain (AI Engineer) https://www.youtube.com/watch?v=CvRngaQZQ3Y

Researchers found a way to extract raw AI reasoning traces, exposing leaked credentials and evidence of model distillation. Meanwhile Grok 4.6 lands near the frontier, Cursor's new multi-agent system lets AI agents message each other, and Claude users are stripping their own instructions to let the model think for itself.

Source episodes:

- Grok 4.6 is Fable now (Wes Roth) https://www.youtube.com/watch?v=VB39bHByHuU

- all AI thoughts JUST got revealed... (Wes Roth) https://www.youtube.com/watch?v=kKjmv2CuVUI

- Grok 4.6 Shows How Fast Your AI Options Are Expanding (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124187865/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-13%2F429795335-44100-2-2cf594557f87b.mp3

- I Deleted All My Claude Skills... And Claude Got Smarter (Nate Herk | AI Automation) https://www.youtube.com/watch?v=XNQBCRcwXV4

- Grok Bot is For Real. What You Need to Know. (Nate Herk | AI Automation) https://www.youtube.com/watch?v=PQBYZQqan2g

- Is Your Grade Book Safe from AI? | Check-In 52 (ChatEDU – The AI & Education Podcast) https://anchor.fm/s/f187ab08/podcast/play/124151961/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-13%2F429746255-44100-2-0f51faa8dbbb9.mp3

- The EASIEST Way to Use Codex and Claude Code Together (Sean Kochel) https://www.youtube.com/watch?v=ZDh1005AOtM

- AI Agents for Performance: Ship Faster, Pay Less — Rajat Shah, Netflix (AI Engineer) https://www.youtube.com/watch?v=CgsWxRUY5Eo

- Microsoft Built a New Charting Language for AI Agents (Better Stack) https://www.youtube.com/watch?v=xFWhiWe91BE

Every frontier model lost money betting on Premier League matches in a new benchmark, while Grok Bot's virtual-coworker launch splits builders between awe and skepticism. Plus: a 24-model coding leaderboard crowns an underdog, and Anthropic's invisible watermarks spark developer backlash.

Source episodes:

- Grok Bot Finally Makes AI Agents Easy (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124140905/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-12%2F429730192-44100-2-8e39630ef7295.mp3

- I Ran 100+ Prompts with 24 LLMs on Go Project (AI Coding Daily) https://www.youtube.com/watch?v=y5Jflajvjps

- Taking Reinforcement Learning Cross Datacenter — Nan Jiang, Modal (AI Engineer) https://www.youtube.com/watch?v=maRzp4kImJ4

- Learning on the Job: The Future of Post-Training — Raymond Feng, Applied Compute (AI Engineer) https://www.youtube.com/watch?v=k35LeKZEhiE

- Scaling to Long Horizons — Ross Taylor & Chengxi Taylor, General Reasoning (AI Engineer) https://www.youtube.com/watch?v=2bvtay8wGYI

- This $8 Microcontroller Is Running a Language Model Locally! (Better Stack) https://www.youtube.com/watch?v=0qXVMt3pIjU

- Agents at Scale: Inside MiniMax's Model and the Infrastructure Behind It — Olive Song (AI Engineer) https://www.youtube.com/watch?v=AVMr9PMINyo

Indy Devdan and Anthropic both split AI agents into a reasoning 'brain' and a sandboxed 'hands.' Anthropic's version cut time-to-first-token 60%. A new routing layer shows cheap local models can replace frontier calls for most tasks.

Source episodes:

- This New AI From SpaceX Is The Future (First Look) (The AI Advantage) https://www.youtube.com/watch?v=ifToIqwU3Jg

- Lessons from the very first chatbot (The Vergecast) https://www.podtrac.com/pts/redirect.mp3/pdst.fm/e/pscrb.fm/rss/p/mgln.ai/e/257/traffic.megaphone.fm/VMP1006809881.mp3

- Engineers… Your Software Factory NEEDS Agent Sandboxes to SCALE (exe.dev) (IndyDevDan) https://www.youtube.com/watch?v=SEI_qIW4o2c

- AI Optimism Has a Trust Problem (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124090320/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-11%2F429659390-44100-2-5bb359a8340ba.mp3

- Evolution of agentic surfaces — Gagan Bhat & Isabella Kai He, Anthropic (AI Engineer) https://www.youtube.com/watch?v=K0X9QDRkIdg

- What is an Agent Router? and How to build a Local One! (Prompt Engineering) https://www.youtube.com/watch?v=ofpqWw0OFKo

- 3D Websites Just Became FREE (One Prompt) (Mark Kashef) https://www.youtube.com/watch?v=FSnubu4Lpz8

- Sergey Brin Retakes Gemini, 4 Labs Lose Containment, Compute Trades at NYSE w/ Kush Bavaria | EP #278 (Moonshots with Peter Diamandis) https://traffic.megaphone.fm/DVVTS1493216813.mp3

- I Tried GPT-5.6-Luna in Cursor vs Codex CLI: Better/Faster/Cheaper? (AI Coding Daily) https://www.youtube.com/watch?v=bFVCcPzQfxU

OpenAI says its new Astro model tested into the critical tier for cyber capabilities, prompting it to delay release and lock down weights. Meanwhile ByteDance is reportedly training a 10-trillion-parameter model despite export controls, and open-weight releases are quietly adopting revenue-sharing licensing.

Source episodes:

- Pencils down: We share our vibe-coded websites (The Vergecast) https://www.podtrac.com/pts/redirect.mp3/pdst.fm/e/pscrb.fm/rss/p/mgln.ai/e/257/traffic.megaphone.fm/VMP8539429926.mp3

- What the Heck is Graph Engineering? (The AI Daily Brief) https://anchor.fm/s/f7cac464/podcast/play/124043095/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-7-10%2F429594264-44100-2-ed4318aa908a2.mp3

- Claude Code and Codex Quietly Learned to Watch Video (Mark Kashef) https://www.youtube.com/watch?v=0I-J1aoxYQY

- Benchmarking Coding Agents on New vs Legacy Codebases — Denys Linkov, Wisedocs (AI Engineer) https://www.youtube.com/watch?v=7vn4WpqNpck

- Anthropic's CCA Exam as a Field-Guide for Agentic Engineering — Frank Coyle, UC Berkeley (AI Engineer) https://www.youtube.com/watch?v=Z-c11pV_uvU

- ChatGPT Codex Voice + browser + Sites: an expert's AI workflow | Nick Baumann (OpenAI) (How I AI) https://anchor.fm/s/1035b1568/podcast/play/123587744/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-6-31%2F428985071-44100-2-37475491155f8.mp3

- MCP Tasks (async): Why Aren't Any Agents Supporting Them? — Cornelia Davis, Temporal (AI Engineer) https://www.youtube.com/watch?v=s4r6nk5WsZw

Sandisk and SK hynix unveiled an open HBF memory spec promising up to 512GB per GPU package for AI inference, but only Google and Tenstorrent have signed on. Red Hat also launched an open source AI governance framework built around EU AI Act compliance.

Source episodes:

- New HBF spec outlines tech that can give GPUs terabytes of extra memory — Sandisk and SK hynix unveil spec with up to 16-Hi NAND stacks, 3 TB/s bandwidth, UCIe — Tom's Hardware (Harold Web Research) https://www.tomshardware.com/pc-components/ssds/sandisk-and-sk-hynix-unveil-hbf-spec-up-to-16-hi-nand-stacks-3-tb-s-bandwidth-ucie

- Red Hat launches new open source project to drive AI governance — IT Pro (Harold Web Research) https://www.itpro.com/software/open-source/red-hat-launches-new-open-source-project-to-drive-ai-governance

- GitHub Trending Today #44: paritok, KiroCrew, wyzer, quickdraw, RunCat Neo, gemma-translator, auteur (Github Awesome) https://www.youtube.com/watch?v=3WpuTNQY5-A

- Velocity Sickness: What Happens When Your Whole Team Gets 10x Faster — Matt Dailey, Ref. (AI Engineer) https://www.youtube.com/watch?v=Kz4QJmNrVXU

- I Tested NEW Muse Spark 1.2 by Meta on 15 Coding Prompts (AI Coding Daily) https://www.youtube.com/watch?v=c-V4MrY03Mc

- Realtime multiplayer, automation, and you! — Idan Gazit, GitHub (AI Engineer) https://www.youtube.com/watch?v=iQ5xldZ9StU

- Guide, Verify, Solve — Anirban Chatterjee, Sonar (AI Engineer) https://www.youtube.com/watch?v=03l29gJXpCE

- Multiplayer agentic engineering — Arjun Singh, Superconductor (AI Engineer) https://www.youtube.com/watch?v=OL7kfezynJM