494: Big Tech Capex, Google & Meta Q1, Tesla Dojo vs Nvidia, Fertility, Framed by AI, and Fraud

"More love."

It's better to be alone than to spend time with toxic people.

It's better to do nothing than to work on something that doesn't matter.

It's better to rest than to climb the wrong mountain.

—James Clear

More love 💚 🥃

🤰🏻🌎👱🏻‍♀️🧒🏼👩🏻‍🦱 Why more humans?

I recently shared a graph showing fertility rates rapidly declining below replacement rate worldwide (more on that later in this Edition), and someone more or less asked me, why is it a bad thing if the human population goes down?

I thought about it for a moment and this is my answer:

I believe that the vast majority of humans are good people with worthwhile lives and they deserve to exist.

I also believe that the planet’s carrying capability is not fixed and that through technology it could be home to tens of billions of people with less pollution than we have now.

I think population size contributes to how vibrant a civilization is. More artists, more geniuses, more entrepreneurs, more explorers, more visionaries, etc. More potential combinations and connections. More love.

That future won’t happen automagically.

It needs to be BUILT.

But it’s possible.

Remember, whatever we do today affects *real* people in the future.

The Universe trying to understand itself through us as sentient beings. You can’t ask for a cooler purpose!

🛀💭 Imagination is stimulation.

Art is the output, packaged in a kind of lossy mental file format that other brains can load up and decode.

It involves putting slices of an inner world into words, music, images, shapes, colors, acting, singing, etc.

We can’t help but see the world through our own subjective perspective. I’m a French-Canadian dude born in the early 1980s, that affects how I see the world.

But I can get a glimpse of other perspectives when I read Le Ly Hayslip’s memoirs about being a woman growing up in rural Vietnam in the 1960s, or when I listen to Beyoncé’s ‘Pretty Hurts’ about the pressures she’s been under since childhood (yea, I know, you didn’t expect me to drop a Beyoncé reference today! I gotta keep you on your toes).

It’s natural to look for perspectives we can easily identify with, like putting on comfy slippers. But if we only seek things that we already know and like, we’re missing out on a lot of growth and discovery.

Your favorite thing is out there, you just haven’t found it yet!

Keep some exploration in your life. Even things that are ultimately not your cup of tea can still teach you something, both about others and about yourself.

⚓️🚢🛳️ I don’t know how, but this chart keeps going!

We passed 23,000 crewmembers on this steamboat. We’ll soon have to either dock in some South Korean shipyard to do structural work and expand, or try to acquire a second-hand cruise ship or black-market aircraft carrier.

We’re getting close to two Hobarts*, which is exciting!

*Context: In July 2020, when I first pushed a raft made of scrap wood in the water, I saw that Byrne Hobart (💚💚💚💚💚 🥃) who writes the excellent The Diff had 12,000 subs. That seemed like an impossibly high number. I thought: If I can ever make it to one Hobart, that would be amazing!

Here we are. Keep going, keep going.

---

🏦 💰 Liberty Capital 💳 💴

💰💰💰🏗️ Big Tech Capex-to-Gross-Profit 🤯

Modest Proposal has this great bit of perspective on just how insane Big Tech is:

Looking at Capex to Gross Profit (to normalize for biz models; I deduct fulfillment expense from AMZN) for GOOG/AMZN/META/MSFT. From 2017 to 2024, Capex/GP is surprisingly only up from 23.2% to 25.8%.

MSFT: 13% -> 26%
GOOG: 20% -> 22%
AMZN: 48% -> 26.7%
META: 19% -> 29.7%

The absolute dollars are astonishing. Capex is going to increase from $48B to $184B. But Gross Profit (ex fulfillment expense) is going to go from $208B to $714B.

This is just bonkers!

Apple’s numbers are lower, but they only seem low *compared* to these others. In absolute terms, Apple spending $10-12bn on capex in the past 12 months is a lot, even if as a percentage of gross profit it’s lower than the rest of the group.

One of the reasons for this is because while they have some of their own data centers, they host a lot of their services on a mix of AWS, Azure, and GCP (ie. iCloud).

It’s probably not a bad idea. No company can be good at everything, and Apple has decided that building a hyperscaler isn’t their area of focus. Because they have so much volume and multi-host, they can probably get very good rates from the big cloud players. They may save some money by building data centers, but that comes at the risk of loss of focus on areas that matter more to their success (ie. product focus).

I recently wrote about how so far the limiting factor on AI has been compute, but if things grow for a while longer, the bottleneck will be energy. Looking at these numbers (almost $200bn in capex), it’s not too hard to imagine that being the case.

I wonder about other bottlenecks, like skilled labor to build and operate gigantic data-centers, or to build and operate the manufacturing plants for all kinds of necessary pieces of the puzzle (a lot of attention goes to TSMC and cutting-edge chips, but there are lots of other parts that go into making servers and networking gear).

📄🔍 Meta and Google Q1 Highlights 💰💰💰🔥