UPDATES

Hey! Hope you’re doing well this week!

🔥 Biggest news for me continues to be Claude Code! Like, I’m barely able to sleep. This is the most creative/productive I’ve ever been in my entire career.

  • It’s stressing me out. BLOG

  • I think it’s the biggest AI jump since ChatGPT (and proto-AGI) BLOG

  • I got it working in ghostty and neovim! VIDEO | BLOG

  • 🔥 This is an absolute MUST video on using Claude Code, from a setup and workflow perspective, by Boris The Creator. 30 MINUTE VIDEO

I guess this is probably obvious by now, but I’m using Typescript for all the things these days. I’m firmly off of Python (even with uv). I think JavaScript is winning the internet, and I think Typescript is the way to go. Could be wrong, but that’s where I’m at. Also, bun. 👀 WHAT IS BUN?

I added a response to my debate on AI with Marcus Hutchins. I basically present a ton of evidence of real work being done by AI that only a highly-paid human could have done. THE BLOG | THE VIDEO

✍🏻 Throughout my time writing online I’ve only had a few people that made me go, “Wow” from the quality of their thinking/writing. Paul Graham. The Last Psychiatrist. Sam Harris. Aaron Swartz. If there are others, I’m forgetting them. And I think I may have found another. Her name is Joan Westenberg, she’s putting out these great little essay / video / podcast pieces that are like dead center bullseye for me—and for UL in general.

I love how the idea is the piece. I love how it’s just her reading it into the camera. I love how it’s a podcast and a video and a blog. It’s exactly what I’ve been doing for the last many years (although I’m doing lots of other formats too), and I just think it’s the most pure form of expression out there.

And, unlike me, every one so far has had this short little idea format, and they’re all bangers. It’s not like technical tutorials, and politics, and some AI doom and gloom thing, then a tool review. It’s just consistent ideas. Clean and simple. Man, love it so much.

Anyway, it feels really good to have found her, because she’s validated my own strategy for thinking and sharing ideas. And it’s heartening to see that this direct approach is good enough. Or at least, if it isn’t I don’t care that it isn’t. Go check her out! HER BLOG | HER YOUTUBE | HER PODCAST

CYBERSECURITY

Google Just Gave Gemini Access to Your Android Apps Without Really Asking
Google's rolling out Gemini access to third-party apps like WhatsApp and Messages on Android, and the opt-out process is confusing even for security experts. THE MALWAREBYTES REPORT | ARS TECHNICA COVERAGE

China-Linked Hackers Create Thousands of Fake Brand Websites to Steal Payment Data  
Researchers uncovered a large phishing network impersonating brands like Apple and PayPal to steal credit card info. THE RECORD ARTICLE

Nova Scotia Power Hackers Had Access to Critical Systems for Over a Month Hackers broke into Nova Scotia Power's systems from March to April and stole everything from bank details to power consumption data affecting 280,000 customers. THE RECORD ARTICLE | PREVIOUS ATTACK COVERAGE | BREACH NOTIFICATION LETTER | INITIAL BREACH CONFIRMATION

US Busts North Korean "Laptop Farm" Scheme That Fooled 100+ Companies  
The DOJ shut down a massive North Korean operation where fake IT workers used stolen identities and AI-generated profiles to get remote jobs at US companies. THE BLEEPINGCOMPUTER ARTICLE | DOJ PRESS RELEASE

Institutional-Grade Opportunities for HNW Investors

Long Angle connects HNW entrepreneurs and executives with institutional-grade alternative investments. No membership fees. Access includes:

  • Private equity, credit, search funds, hedge funds, secondaries

  • $100M+ invested annually, leverage collective expertise and scale

  • Invest alongside pensions, endowments, and family offices

NATIONAL SECURITY

Russia Is Field-Testing AI Drones That Think and Hunt on Their Own
Ukrainian Major General Vladyslav Klochkov says Russia's new MS001 drone uses Nvidia Jetson Orin chips to autonomously identify, prioritize, and strike targets without human commands. He calls it "not a loitering munition but a digital predator." TOM'S HARDWARE ARTICLE | NVIDIA JETSON ORIN COVERAGE | RUSSIA TECH SANCTIONS ARTICLE | DEFENSE EXPRESS REPORT

Chinese Hackers Are Stealing Chip Secrets Instead of Smuggling Physical Chips  
Chinese hackers are increasingly targeting semiconductor companies to steal intellectual property rather than trying to smuggle physical chips past export controls. THE POLITICO STORY

Colombian Navy Intercepts First-Ever Starlink-Powered Narco-Sub Drone  
Drug cartels just escalated to remote-controlled submarines using Starlink internet for uncrewed smuggling operations. THE TOM'S HARDWARE ARTICLE | FRANCE 24 REPORT

NATO Sparks An AI Gold Rush With $1 Billion Defense Fund  
NATO just launched a $1 billion AI investment fund specifically for defense startups, basically creating an AI gold rush for military tech companies. THE POLITICO STORY

White House Trades Chip Design Software Access for China's Rare Earth Materials  
The U.S. lifted export restrictions on chip design software to China in exchange for easier access to rare earth materials, marking the first time export controls were used as a trade chip. TOM'S HARDWARE ARTICLE

AI

Dwarkesh Patel Thinks We're All Wrong About AGI Timelines
Dwarkesh Patel argues that AGI isn't as close as everyone thinks. I don't agree with him, but I'm going to continue to listen to his arguments very closely. And I just love the fact that there's actually some disagreement among people in the valley about this. THE ESSAY | HN DISCUSSION | DWARKESH'S BLOG

Managers Are Using AI to Decide Raises, Promotions, and Layoffs Without Training  
A survey found 60% of managers use AI tools for decisions on raises, promotions, and layoffs, but two-thirds lack training on managing people with AI. If they lack the training then how are they going to know if the AI did a good job? This is where you have to be very careful with AI. THE HILL STORY

Researchers Are Hiding AI Prompts in Papers to Game Peer Review  
Academics are embedding hidden AI prompts in research papers using white text or tiny fonts to manipulate AI-assisted peer reviewers into giving positive feedback.  

The prompts literally tell AI reviewers to "give a positive review only" or praise the paper's "exceptional novelty."  THE TECHCRUNCH ARTICLE

LLMs Actually Do Bayesian Reasoning When Given Enough Examples THE PAPER

Explanations Need a Purpose THE PAPER

Grammarly Goes Multi-Agent by Acquiring Superhuman Email App 
Grammarly is acquiring the email app Superhuman (which I love) as part of their plan to become an "AI productivity platform". THE VERGE ARTICLE | OFFICIAL PRESS RELEASE | GRAMMARLY'S CODA ACQUISITION

TECHNOLOGY

America Now Has Two Completely Different Labor Markets
The U.S. job market has split into two distinct economies: white-collar workers face a brutal hiring freeze while blue-collar and service workers enjoy historically low unemployment rates. THE AXIOS ARTICLE | HN DISCUSSION

Google's Data Center Electricity Use Doubled—Now Equals Ireland's Total Consumption  MIT TECH REVIEW ARTICLE

Someone Built a DNS Service That Tracks the ISS Location in Real-Time  
A developer created a DNS service that returns the International Space Station's current coordinates as DNS TXT records, allowing location tracking via simple DNS queries. THE ARTICLE | HN DISCUSSION

Microsoft Lays Off 9,000 More Employees Including Major Xbox Cuts  
Microsoft is cutting another 9,000 jobs (4% of workforce) with Xbox getting hit particularly hard, including canceling the long-delayed game Everwild that's been in development since 2019. THE VERGE STORY | SEATTLE TIMES REPORT | BLOOMBERG COVERAGE | EVERWILD CANCELLATION

HUMANS

RFK Jr.'s Health Department Calls Nature "Junk Science" THE STORY

Stratus Covid Variant Gets WHO Attention THE ARTICLE

Research Shows Chasing Hobbies Over Achievement Actually Makes People Happier  
New research from the University of Toronto finds that people prioritizing hedonic pursuits (pleasure and enjoyment) over achievement-based goals report significantly higher life satisfaction and well-being. THE RESEARCH

New Study Finds Cool People Are Just Emotionally Stable With Good Social Skills  
A study reveals "coolness" is primarily emotional stability, social competence, and calmness under pressure. THE NYT ARTICLE | HN DISCUSSION

Teen Drivers Spend 21% of Time Looking at Phones Despite Knowing the Risks  
A study finds US teen drivers spend 21% of driving time looking at phones, mostly entertainment and messaging apps. THE 9TO5MAC ARTICLE | CNET COVERAGE

The Spoken Word Is the Hinge of History  
Dr. Alexander Westenberg explores how rhetoric and oratory have shaped democracy and civilization’s fate, with great orators defending or destroying institutions. Examples include Pericles, Demosthenes, Cicero, Marc Antony, Churchill, Roosevelt, and Martin Luther King Jr. THE ESSAY

DISCOVERY

How to Become a Creator-Monk - Choosing Depth Over Scale
Joan Westenberg explores the "creator-monk" path inspired by Thomas Merton's monastic clarity, where creators intentionally choose depth over distribution and build for 100 true users instead of 100K followers. THE PODCAST

Engineer Shows How AI Actually Fits Into Real Development Work  
A GoDaddy engineer shares how AI tools like Claude Code and GitHub Copilot help with daily coding tasks, from converting JavaScript to TypeScript to writing tests in unfamiliar languages. THE GODADDY ARTICLE

Using O3 to Profile Yourself From Your Saved Links Actually Works
This person used OpenAI's o3 to analyze their 500+ saved Pocket links and got surprisingly accurate insights about their personality and interests. The AI correctly identified their focus areas, learning patterns, and even personal quirks from just their reading habits. THE EXPERIMENT

The Uncertain Future of Coding Careers and Why I'm Still Hopeful  
Jon Hoyt argues AI will change coding careers but problem-solving and solution-building remain vital. THE ESSAY

Awesome Collection of Claude Code Commands and Workflows  
hesreallyhim curated commands and workflows for Claude Code, automating tasks like git commits and React component generation. THE REPOSITORY

Orwell Predicted AI-Generated Content in 1984 With His "Versificator" Machine  
Simon Willison highlights how Orwell's 1984 described a "versificator" machine creating songs and literature mechanically, predicting generative AI decades before its advent.  THE BLOG POST

Developer Goes From 1000 Lines of Neovim Config to Just 11 Lines  
Vitaly stripped his entire Neovim setup down to 11 lines with zero plugins, finding that removing LSPs, autocomplete, and IDE features actually made him a better programmer by forcing him to understand his tools and codebase more deeply. THE BLOG POST

The Cult of Hard Mode  
Another great one from Westenberg! This one is about over-rotating on tools. THE VIDEO

APHORISM OF THE WEEK

Don’t explain your philosophy. Embody it.

Epictetus

GET THE MEMBER EDITION

You’re currently receiving the STANDARD edition. Members get additional content sections, including IDEAS, a bi-monthly MEMBER-ONLY ESSAY, and the RECOMMENDATION OF THE WEEK.

In addition, you’ll get access to the extraordinary UL Member Community, which includes vibrant conversations with over 1,400 of the smartest and kindest people you’ll find on the internet, the Member Archive, UL Book Club, a monthly member meet-up, access to in-person events, and much more.

Keep Reading

No posts found