512: Crowdstrike's Very Bad SNAFU, Meta vs EU, Anduril & SpaceX, OpenAI's GPT-4o Mini, Sam Walton, LLMs + IRS, and Karate Kid

"the 4th anniversary of this newsletter"

Those who lack the courage will always find a philosophy to justify it.

–Albert Camus

Celebrate good times, come on!

⚓️🛳️ ✉️📬 🎂Two days ago was the 4th anniversary of this newsletter.

I started it on a whim on July 20, 2020 (coincidentally, the 51st anniversary of Apollo 11 touching down on the moon! 🚀🌙). I needed a new hobby, a way to share the stuff I found interesting and random shower thoughts. 🛀💭

I never expected anything from it. But over 500 Editions + 35 podcasts and interviews, it has grown into one of the most rewarding things in my life, allowing me to meet many great people — some of whom became close friends. It also refined my thinking, because writing is thinking, and you often don't know what you think until you've put it down on the page.

(That’s why I highly recommend that you start writing in public. It doesn’t matter if anyone reads it or not — don’t get me wrong, it’s a very nice bonus when they do — but just the fact that your stuff is out there and that someone could read it will force you to think about things differently.)

There were no big spikes or viral moments, just chugging along week after week. 🚂

Today, over 24,000 people are receiving these things in their inboxes when I go clickety-clack on the keyboard and press a button, which seems nuts to me. That's like a stadium crowd 😳

I really appreciate anyone who reads this. I know that you could be doing something else with your time, like scrolling Instagram to watch fitness influencers do heavy deadlifts. I never take it for granted.

Here's to the next 4, let's see where it goes 💚 🥃

🛀💭 What jealous people need to understand is that they’re not jealous of a specific thing.

If they got the thing they want they would eventually transfer that jealous energy to something else.

Chronic jealousy is a generalized condition. It’s not about anything specific. It's often a symptom of deeper insecurities or dissatisfaction with one's life.

The best way to cure it is to take action to improve things that are meaningful to you — bring back to the focus to what you control and and stop focusing so much on others.

💰💸📈📉🔪🥧 Most funds have lots of positions. 30+ seems to be common. Some have significantly more.

When you are invested in that many things, chances are you can always talk about a few investments that are doing spectacularly well lately.

But if you split the pie into 30 or 50 slices, even if you always have something to brag about by cherry-picking, the winners may not move the needle much.

🏛️⚔️ The Gallic Wars series on Caesar’s “masterwork of psychology, strategy, and propaganda” continues. Part 3 is out!

• 🎧 Caesar At Sea — De Bello Gallico - Part 3

---

🏦 💰 Liberty Capital 💳 💴

🦙✋🇪🇺 Following Apple’s lead, Meta won't offer future multimodal AI models in the EU 🍎🚫🤖

Twice is a trend:

“We will release a multimodal Llama model over the coming months, but not in the EU due to the unpredictable nature of the European regulatory environment,” Meta spokesperson Kate McLaughlin said [...]

The United Kingdom has a nearly identical law to GDPR, but Meta says it isn't seeing the same level of regulatory uncertainty and plans to launch its new model for U.K. users. (Source)

Will Google be next? OpenAI/Microsoft? Anthropic? Nvidia?

EU regulators probably feel pretty good about their odds when taking on these companies separately and as long as EU citizens aren’t impacted by what is going on, but if a bunch of tech companies all stop releasing various products and services in the EU at the same time, it could put real pressure on regulators to rethink their approach.

🤖🛡️⚔️ Anduril and SpaceX: The young companies challenging the old US Defense Primes 🛰️🛰️🛰️

The graph above shows the market cap of the biggest publicly traded U.S. defense companies (the surviving ones, as there was significant consolidation after the end of the Cold War).

It took them decades to get where they are:

• RTX (aka Raytheon) has a market cap of about $135bn

• Lockheed Martin around $115bn

• Boeing (who also has a large civilian presence) is around 110, but as you can see, they peaked around $250bn

• General Dynamics around $80bn

• Northrop Grumman is around $65bn

In the other hand, Anduril was founded in 2017. Its most famous co-founder, Palmer Luckey, was born in 1992, soon after the collapse of the USSR.

In that short period, the startup has positioned itself as a fast-moving, less bureaucratic alternative to the prime contractors, better able to supply the latest tech for the battlefield, including artificial intelligence and robotics.

Anduril's major products fit into a few categories: unmanned aerial systems (UAS), counter-UAS (CUAS), semi-portable autonomous surveillance systems, and networked command and control software. Rather than do a cost-plus procurement process which incentivizes spending a lot of time and money on every weapons system, they tend to develop things on spec, with their own R&D dollars, and keep most of their IP.