523: Success is Not Enough, Crowdstrike, AMD, Oracle AI Capex, Meta's 100k GPUs Cluster, China's VCs, OpenAI o1 Complex Reasoning Model, and LEGO Movie

"when your output needs to be graphed on a log chart"

Slowing down enables you to act in a high quality way.

Kind rather than curt. Polished rather than sloppy.

It's hard to be thoughtful when you're in a rush.

–James Clear

🪄 Everyday magic 🌈

🌈🦄🚰 🔌⚡️💡 In the category of ‘not taking for granted the magical infrastructure around us’, Midjourney Founder David Holz had a good riff:

[I] had an emotional moment staring at a running faucet this week. we really don't appreciate civilizational infrastructure. literally "pay as you go" unlimited clean drinkable water delivered to everyone, physically piped straight to you and whisked away as soon as you dont need it

Friend-of-the-show Mark Nelson (check out the podcasts we did together) expanded that thought:

The grid delivers you exactly as much energy as you call for exactly when you ask for it straight into your house from giant electromagnetic force field generators up to hundreds of miles away with almost no losses along the way

We shouldn’t take that stuff for granted!

🎶🎻🎼 I was thinking about modern “classical” and orchestral music composers. Since they mostly earn a living making film scores, the non-musical elements of a film have a big impact on their careers.

If they compose amazing music but the film sucks, few people will ever hear it and it will quickly fade into obscurity.

If they write merely okay music but the film is *amazing*, the association with it can create a halo effect and give the music longevity and recognition. It will also ensure they get more work and bigger fees.

Ideally, great music is paired with a great film (*cough* John Williams Hans Zimmer Ennio Morricone Jerry Goldsmith James Horner *cough* wow that was a big coughing fit) but that’s largely out of most composers’ control.

🤔 🕰️ 📲 ✋🇨🇳 It's been a while since I've heard anyone talking about banning TikTok or forcing a spin-off to a non-Chinese entity.

What happened behind the scenes to change the discourse? I know there are court cases, but that has never stopped politicians before, especially in an election year.

Maybe it’s because the platform is so popular that nobody wants to be seen as attacking it too much — they’re afraid of the backlash..?

I think the way around that is to be clear about forcing a spin-off that isn’t controlled by China, rather than imposing an outright ban. It almost happened once (remember the craziness with Microsoft, Walmart, and Oracle?).

Why not finish the job and remove the CCP psyops vector?

☕️😩⚙️ In the ‘Well, this is annoying’ category, my Fellow Opus coffee grinder has let me down.

I’m not sure what happened, but less than a month after I bought it, it suddenly started clogging up and had tons of retention. The first time it happened, it seemed like a magic trick — I’d put beans in, grind them, and nothing would come out the bottom.

I realized that the entire interior of the chute was packed tightly. It took me a while to clean and empty it, but even after that it, it remained unreliable and I never knew if every gram I put in would come out or just 50%...

Maybe the ionizer got screwed up, and rather than reduce static, it started creating some. ¯\_(ツ)_/¯

I’ve lost trust in that design, so I'll probably go with something else now. I’m just not sure what yet 🤔

☎️🗣️🗣️🗣️🗣️🗣️ On Wednesday, we had the second Zoom Q&A video call for paid supporters, and it was GREAT!

A very big MERCI to all who joined, I really enjoyed it. It was a ton of fun and I’m looking forward to doing it again soon 💚 🥃

---

🏦 💰 Liberty Capital 💳 💴

Moderna, Zoom, and Pfizer since mid 2020.

🫣 Success is not Enough: Capitalism is Brutal 😫

You have a breakthrough product that the whole world suddenly needs. Your revenue grows by hundreds of percent and your free cash flow explodes to the moon…

…and yet, a few years later your stock might be flat or barely higher (see above for Moderna, Zoom, and Pfizer, but there are countless others).

Pfizer’s 10 and 20-year stock CAGR is around 4%, and it has been underperforming the S&P500 over decades. This company is known for Xanax, Zoloft, Lipitor, Viagra, Advil, and COVID19 mRNA vaccines — many blockbusters and household names generating billions! Yet they couldn’t outperform…

Long-term success in the stock market is *hard*.

In competitive sectors of the economy (which isn’t all of them, to be clear!), most of the value created *isn’t* captured and society as a whole benefits from large positive externalities. In other words, if you live in a market economy, there are millions of people working very hard mostly for your benefit.

The very large and very profitable companies at the top of the power law are extremely visible, but there are millions of companies just in the US, and for most of them, it’s a knife fight just to stay in the game and make a bit of profit… Even just among very large publicly traded companies, a small group provides most of the return of the S&P500 over time and most of the rest underperform.

As I said:

💃 AMD Sets Faster Tempo to Keep Up with Nvidia 🏗️🤖

The semiconductor industry has never been slow — when your output needs to be graphed on a log chart, you’re pretty productive — but in recent times, Jensen has been pushing a pretty incredible tempo. ⏱️

By definition, if you can’t keep up, you fall behind (though the fast tempo does come at a price and increases the possibilities of a misstep… 👣💥).

Lisa Su knows how to tango: