Hey debriefer!
The Pope just sat next to an Anthropic co-founder, wrote a 42,000-word warning about AI, and called for the whole industry to slow down. Meanwhile, Anthropic is closing a $900 billion round - bigger than OpenAI. OpenAI's reasoning model just disproved an 80-year-old math problem on its own. And I built a 7-agent AI content team in Claude Code that runs my entire YouTube pipeline.
This week you can build one too.
Let's debrief.
In today's AI Debrief:
📰 The Pope's first encyclical: AI on trial at the Vatican
💰 Anthropic closing a $900B round (now bigger than OpenAI)
🧮 OpenAI's model solved an 80-year-old math problem solo
⚡ Cursor Composer 2.5: Opus-level coding at 1/10th the cost
🔧 Tool of the Week: Build your own AI content team in Claude Code
🎓 Skill of the Week: How I built a 50K newsletter ($3K/month)
🏠 Inside Try AI Academy: Jarvis, Apple giveaways, members helping members
THE VATICAN VS THE VALLEY
POPE LEO PUT AI ON TRIAL - AND BROUGHT ANTHROPIC WITH HIM
Image Source: ncronline.org
The debrief: On Monday, Pope Leo XIV released his first encyclical - a 42,000-word document called "Magnifica Humanitas" - calling for the "disarming" of artificial intelligence. Sitting next to him on stage: Anthropic co-founder Chris Olah.
The details: This is the first time a sitting pope has personally presented an encyclical to the world, and the first major Catholic teaching document specifically about AI. Leo signed it on May 15 - 135 years to the day after Leo XIII signed Rerum Novarum, the Industrial Revolution doctrine on workers' rights. The pope warned against the "race for ever more powerful algorithms and larger datasets, driven by the desire to secure geopolitical or commercial dominance," and urged governments to slow AI development. Olah didn't play defense - he conceded that every frontier AI lab "operates inside a set of incentives and constraints that can sometimes conflict with doing the right thing," and called on religious communities, civil society, and governments to push the technology toward better outcomes. The optics weren't subtle: Anthropic is already suing the Trump administration over its refusal to allow the U.S. military unrestricted use of Claude, and Leo's encyclical doubles down on monitoring AI in warfare.
Why it matters: The most powerful religious institution in the world just framed AI as the defining moral test of our generation - and aligned with the lab that's positioning itself as the ethics-first alternative to OpenAI. For builders, this isn't background noise. It's the start of a regulatory and reputational divide: labs that lean into safety and constraints will keep winning institutional trust (and government contracts where ethics matter), while labs racing to ship will catch the headwinds. If you're building on top of these models, factor the politics into your stack now - the model you pick is a brand statement, not just a performance one.
💰 Anthropic is closing a $900 billion round - making it the most valuable AI startup on earth. Bloomberg reports Anthropic is wrapping a $30B+ raise this week led by Sequoia, Dragoneer, Altimeter, and Greenoaks. The valuation jump is brutal: $380B in February → $900B+ now. Anthropic is projecting Q2 revenue of $10.9B and a $50B annualized run rate by end of June. Three months ago OpenAI ($852B) was the king of AI valuations. This week, Anthropic took the crown.
🧮 OpenAI's reasoning model solved an 80-year-old math problem on its own. An unreleased general-purpose reasoning model disproved the Erdős unit distance conjecture - a discrete geometry problem first posed in 1946 - by producing a 125-page proof from a single open-ended prompt. Nine independent mathematicians, including Fields Medalist Tim Gowers, verified it. OpenAI is calling it "the first time AI has autonomously solved a prominent open problem central to a field of mathematics." Translation: AI just stopped being a tool for math and started being a contributor to it.
⚡ Cursor dropped Composer 2.5 - a new coding model that matches Claude Opus 4.7 and GPT-5.5 on benchmarks at roughly 1/10th the cost. It scores 79.8% on SWE-Bench Multilingual (Opus is 80.5%) and costs about $0.50 per task vs. $7 per task for Opus. It's only available inside Cursor, but the signal is bigger than the product: every IDE that doesn't ship its own model is about to feel the squeeze.
🍎 Apple's Siri overhaul is finally real. Bloomberg got the first leaked screenshots of iOS 27 today, with a fully redesigned Siri and a new chatbot-style app as the centerpiece. Apple announces officially at WWDC on June 8. After 18 months of "Apple is behind on AI" takes, Cupertino is about to make its first real swing - and 1 billion+ iPhones will be the test bed.
Big AI just consolidated more power this week - and even the Pope called it out. But you don't have to wait for the giants to share the leverage. Build your own AI team in Claude Code this weekend: seven agents that find trending topics, write scripts, design thumbnails, and run your entire content pipeline while you sleep.
BUILD YOUR OWN AI CONTENT TEAM IN CLAUDE CODE
I just built a 7-agent system inside Claude Code that runs my YouTube pipeline end-to-end. Channel analyst studies my voice. Trend scout finds outlier videos every morning. Script writer drafts hooks in my tone. Thumbnail designer generates 3 concepts with my face composited in. Video editor cuts the raw recording. Daily reporter emails me a brief with everything ready to record.
Why this matters now: Anthropic just hit $900B. The big labs are racing to build agents you'll rent for $100/month. Claude Code lets you build your own today - one project, seven specialized sub-agents, sharing state through Google Sheets. The ROI math is brutal: this stack costs ~$75-85/month and replaces a freelance social media manager, editor, and researcher who'd charge $4-7K combined.
The 3-step start:
Install Claude Code (
npm install -g @anthropic-ai/claude-code), grab your Anthropic API key, and create a single project folder with aCLAUDE.mdfile - this is the brain every agent reads firstBuild the first loop only: Channel Analyst → Trend Scout → Script Writer → Daily Reporter. Don't try to build all 7 on day one. Get the idea-to-script-to-email loop working first
Layer in thumbnails, video editing, and shorts repurposing in week 2 once the core loop is humming
👇 Watch the YouTube walkthrough below
Stuck applying this? 2,000+ readers inside The AI Academy are figuring it out together. Join free →
HOW I BUILT A 50,000 AI NEWSLETTER ($3,000/MO)
This week, learn the exact 3-layer stack that took The AI Debrief from zero to 50,000+ subscribers in 60 days - generating $3K/month while I sleep. No team. No ad budget beyond the engine itself. Just three tools wired together correctly.
Most newsletters die at 1,000 subs because creators only build Layer 3 (writing). Build Layer 1 and 2 first.
The 5-step build:
Layer 1 - Distribution: Turn on Beehiiv Boosts. Pay $1.50-$3.50 per confirmed subscriber. Start at $50/day, scale as ROI hits. This is the #1 growth lever and most creators skip it for "real" subscribers - Boosts subscribers ARE real
Layer 2 - Conversion: Build a ManyChat IG DM funnel. Post a video, ask viewers to comment a keyword, auto-DM them the signup link. Beats a bio link 5-10x because the platform rewards comments and the DM feels personal
Layer 3 - Content Engine: Build a Claude Project loaded with your prior issues, audience profile, and brand voice. Spend 4 hours upfront training it - this is the part everyone skips. Output: drafts that need 10 minutes of editing, not 2 hours of writing
Pick a tighter niche than feels comfortable. "AI" is too broad. "AI for marketers" or "AI for VCs" converts better on Boosts, opens higher, and attracts better sponsors
Monetize on the floor first: Beehiiv ad network + Boosts on the receiving end = ~$2K passive baseline. Direct sponsors are the variable upside. Don't chase sponsors until you have 10K+ engaged subs and a media kit
📘 Grab the full playbook (Claude Project setup, ManyChat templates, Beehiiv Boosts categories) inside Try AI Academy (free) →
👇 Watch the full walkthrough below
Reading is the warm-up. Building is the workout. Join 2,000+ readers inside The AI Academy →
INSIDE THE AI ACADEMY
A great thread fired off this week: member Marwen B. asked "What's the relationship between Jarvis and Claude?" - kicking off a real conversation about local AI, APIs, and how to build your own assistant on top of Claude's brain. This is what's working inside the community: people walking through the guides together, breaking down what each tool actually does, and helping each other build instead of just consume.
Plus - monthly Apple giveaways are still running. Active members get entered every month to win Apple gear (AirPods, iPads, MacBooks). Free to join, free to enter.
EVERYTHING IN ONE PLACE
🧰 The AI Academy - our free AI community with over 2,000 members
📚 Archive - every past edition in one place
✍️ Write for Us - share your AI expertise with 100,000+ readers
💸 Referrals - refer friends, earn real rewards
📣 Advertise - reach 100,000+ AI professionals
📣 Contact Us - get in touch with us for anything
Share The AI Debrief. Get paid.
Refer 1 → your name shouted out to 100,000 readers
Refer 5 → mystery reward with real dollar value
Refer 10 → $100 back, every time. No cap. Forever.
This week's shoutouts:
Pieter V. · Saanvi R. · Tobias M. · Anuk D. · Felipe O. · Kenji H. · Mariella B. · Tomek K. · Yuki S. · Idris N. Pieter V. · Saanvi R. · Tobias M. · Anuk D. · Felipe O. · Kenji H. · Mariella B. · Tomek K. · Yuki S. · Idris N. · Caleb F. · Priya M. · Diego R. · Aanya V. · Ezekiel T. · Linnea J. · Rashid A. · Zaina H. · Soren P. · Naveen S. · Catalina M. · Bohdan Z. · Amaia L. · Theo G. · Wren E. · Olu B. · Mihail D. · Sienna O. · Yara K. · Kofi A. · Romina C. · Esteban N. · Anika T. · Ravi S. · Mateusz J. · Imani O. · Lucca F. · Nadya R. · Bilal A. · Iker S. · Camille D. · Olamide J. · Tomas R. · Cyrus M. · Avery K. · Roshni B. · Magnus E. · Selene V. · Tariq H. · Eliana P.
Want to see your name here next week? Refer one friend with your unique link below 👇
NEXT FRIDAY…
Apple announces the new Siri at WWDC on June 8. I'll break down what builders should actually pay attention to - and the free Apple alternative most people are sleeping on right now.
Thanks for reading. Our mission is to educate as many people as possible around AI literacy - see you next week.
- Drew & the rest of the humans behind The AI Debrief
