336: Could Apple Get Out of China?, Elon Musk + Twitter, ASML Documentary, Economics of 3D Assets, Micron's $100bn, and SR-71’s P&W J58 Engines
"alone with my thoughts and even *gasp* feel boredom"
Keep ignoring feedback and life will keep teaching you the same lesson. —James Clear
☑ 🗳👦🏻 On Monday, I voted in a provincial election. I brought my 8yo to show him the process and get him used to the idea that this is an important privilege and duty.
I was pleasantly surprised to see that they were set up for kids.
They gave him a “young voter” type of sticker, they had a booth where he could do a mock ‘vote’ with a ballot targeted to kids, and they even gave him a temporary “I voted” tattoo.
Very cool. Maybe it has always been like this and I just never noticed because I didn’t go with a kid (or maybe they don’t do it when polling stations are very busy), but I think every electoral location should do things like this.
🏃♂️👟 📊 Yesterday, I ran for the 40th time since I started running on May 20, 2022.
Every previous run, I had done the same loop, but this time I decided to expand it by about 25%. I expected it to be really difficult, but it was fine, showing me once again that a lot of perceived barriers in this sport are psychological, and I should probably ‘just go for it’ earlier.
Worst case, I have to slow down or walk part of it, but I’d still progress faster than if I hadn’t pushed myself.
You know I’m a nerd, so of course I track metrics every time I run. This is my pace (in min/km) since I started:
Getting better. I can also see correlations between really hot & humid days or bad nights of sleep, etc, and slower pace.
☢️ 🔌 Friend-of-the-show and supporter Brent Muio (💚 🥃) has a good thread about David Tepper buying a 10% position in Constellation Energy.
🥵🥱 Great podcast about ‘the comfort crisis’:
Makes me want to hike with a weighted rucksack without listening to anything, so I can be alone with my thoughts and even *gasp* feel boredom…
Michael and Peter quite like the backpacks made by Goruck.
It does make intuitive sense that this would be very effective considering that every military in the world has been using it forever to whip recruits into shape.
I’ll probably start just by putting a few textbooks in my regular backpack and hike in the woods to see how I like it 🤔 🌲🌲🌲🚶🏻♂️🌲🌲🐿🌲 (yea, there’s always giant squirrels)
The concept of ‘misogi’ that is discussed is also very intriguing.
The general idea is basically to, at least once a year, really push yourself to discover your true ability/limits.
Take on challenges that radically expand your sense of what’s possible. There are just two rules: you have a fifty-per-cent chance of success at best, and it doesn’t kill you…
a misogi is a physical trial that you don’t practice or prepare for (no marathons), you don’t perform before a crowd (no CrossFit-style competitions) [and you can’t post on social media], and you don’t brag or pay to enter (no Tough Mudders). Thinking outside the box is important, too.
In the modern world, we so rarely find out where the edges of our capabilities are (mentally, and physically) that it can be quite useful to do it on-purpose, both for personal growth and to create some perspective with life’s regular trials.
🇺🇦🇷🇺 Garry Kasparov has a good op-ed about Russia’s war.
💚 🥃 🥦 I know you’ve been sitting on the fence for a while, meaning to support this project but always putting it off to a later time because nothing is forcing you to and nobody likes paying for things.
But your support makes a big difference and allows me to keep going and do more. Thank you for doing it, your kindness is appreciated 🧡
A Word From Our Sponsor: 📈 Revealera 📊
Revealera provides data and insights for investors into hiring trends for 3,500+ public/private companies + technology popularity trends for 500+ SaaS/Cloud Products.
We give investors insights into:
Job Openings trends: Insights into a company’s growth prospects.
Technology Popularity Trends: Insights into how widely products like Datadog, AWS, Splunk, etc, are gaining adoption.
Vendor Sign-ups (Currently Alpha) tracks the # of companies, as well as the specific companies, that have signed up for SaaS products such as Zoom in near real-time.
Visit Revealera.com for a ✨free✨ trial/demo.
🏦 💰 Liberty Capital 💳 💴
🍎🇨🇳 ‘it would take about eight years to move just 10% of Apple’s production capacity out of China’
Bloomberg Intelligence estimates it would take about eight years to move just 10% of Apple’s production capacity out of China, where roughly 98% of the company’s iPhones have been made. Scores of local component suppliers — not to mention modern and efficient transport, communication and electricity supplies — make it particularly difficult to get out of the world’s second-largest economy.
“With China accounting for 70% of global smartphone manufacturing and leading Chinese vendors accounting for nearly half of global shipments, the region has a well-developed supply chain, which will be tough to replicate -- and one Apple could lose access to if it moves,” BI’s report from analysts Steven Tseng and Woo Jin Ho said.
In edition #333, I wrote about Apple moving some manufacturing/assembly to India. We should keep in mind that there’s a difference between ‘final assembly’ and the manufacturing of the parts. I wrote:
Some may call it “manufacturing the iPhone in India”, but things are a bit more complex than that…
iPhone parts are manufactured all over, some in Japan, some South-Korea, some China, some the US, etc. Final assembly is what we’re talking about.
Most of the value of all this certainly flows to the US, where the intellectual property largely originates…True, true.
But assembly matters a lot! And the manufacuring of low-value parts may still be very important *and* hard to rapidly move elsewhere.
There’s probably a bunch of tiny screws and other bits that are made with very high precision and tolerance, with local manufacturing hubs having years and years of experience making them right, toolmakers and machinists in numbers to do Apple-scale (which is pretty unique). If you don’t have that part, it’s still a show-stopper even if it’s not as important as the semiconductors parts…
Not easy to spin all that up somewhere else…
Bloomberg Intelligence says that overall tech-industry dependence could be reduced by 20%-40% “in most cases” by 2030. For hardware and electronic manufacturers, they could reduce their reliance on the Chinese market to 20%-30% over the next decade, BI calculates.
Who knows if any of these predictions are anywhere near correct, and whether there’s path-dependency that could mean that some of the more extreme (and seemingly unlikely at first glance) scenarios may not actually be pretty likely 🤔 AAPL 0.00%↑
🐦 Looks like Elon Musk will buy Twitter after all 💸
Not that he didn’t try to get out of it any way he could think of, but in the end, it looks like this may be happening (notice all the conditional clauses — don’t we all feel like the cat that once sat on the stove? 🐈⬛):
Elon Musk is proposing to buy Twitter Inc. for the original offer price of $54.20 a share, Bloomberg News reports.
Musk made the proposal in a letter to Twitter, according to people familiar with the matter, who asked not to be identified discussing confidential information. Shares in Twitter climbed as much as 18% on the news, after trading was briefly halted.
Now the hard part. Now comes the much wider band of possible outcomes for the service that we all love-and-love-to-hate so much. TWTR 0.00%↑
Hold on to your butts, this may be a wild ride 🎢 😬
More thoughts from friend-of-the-show and supporter (💚 🥃) Andrew Walker, who I’m very happy for:
The economics of 3D asset production (for video games, films, metaverse, etc) 🎮 🎥 💻 🤖 + Discontinuity
It’s a good thing that AAA video games can make a lot of money, because they are ridiculously expensive to create (at least to AAA standards of visual pyrotechnics).
This cost is largely for armies of people who make the digital assets used to build 3D worlds.
Everything needs to have textures, so you’ve got people making those. Then you have to pull these textures together over 3D models, so you’ve got lots of people making those, from a rock by the side of the road to very complex character models. Then anything that moves must be animated, and while you can do some mo-cap or find some shortcuts, a lot of it is still done by hand, or at least, overseen/corrected by humans to look good.
We’re just getting started!
More people take these 3D assets and build environments with (from single rooms to multi-story buildings to whole cities and countries!), then more people have to work on story, gameplay (beta-testers to make sure things are fun and balanced), voice-acting, music, lighting, in-game weather patterns! etc. I’m probably forgetting a bunch of stuff — but you get the idea.
Very expensive. And as computers and GPUs get better, resolutions are going up, and all these textures and 3D assets need to be created in higher-fidelity to look good, which requires even more work. The advent of real-time ray-tracing also means that fewer artistic shortcuts can be used with light-maps, etc.
Are costs going to keep ballooning up, or will some discontinuity break the cycle?
AI is having a real moment right now, and I hate to use it as the answer to everything (it’s better when the answer is surprising, right?), but this is another area where it seems like the math only works if a lot of 3D assets can be generated automatically, or at least extracted from other existing assets that don’t have to be created de novo at great expense.
For example, Nvidia showed how they could take 2D photos of various objects, animals, plants, vehicles, whatever, and turn them automagically into 3D models.
Trained using only 2D images, NVIDIA GET3D generates 3D shapes with high-fidelity textures and complex geometric details. These 3D objects are created in the same format used by popular graphics software applications, allowing users to immediately import their shapes into 3D renderers and game engines for further editing.
Once you start to mix-and-match these tools and daisy chain them, you can probably get pretty wild results. We know that generative text-to-image models like Stable Diffusion can create all kinds of unique variations.
Imagine easily turning some of the best ones into 3D assets for a game or to be included in the metaverse, and then animating it (if it’s something that move) using a similar technique that has learned how certain things tend to move...
Procedurally-generated game/metaverse environments — think Minecraft but without the need to make everything blocky. Imagine Call of Duty graphics, but a much larger, dynamic world, or a new world every time you restart the server (not being limited to just a few hand-crafted maps). This is coming.
‘Micron Pledges Up to $100 Billion for Semiconductor Factory in New York’ 💰💰💰💰💰💰🗽
Mr. Moneybags is coming to town! (I mean, they always use numbers over decades to have the biggest amount possible, but it’s still big bucks)
Micron announced on Tuesday that it planned to spend as much as $100 billion over the next 20 years or more to build a huge computer chip factory complex in upstate New York, the latest move by a major semiconductor maker to invest in the United States. [...]
“There is no doubt that without the CHIPS act, we would not be here today,” said Sanjay Mehrotra, chief executive of Micron.
The legislation along with favorable tax treatment and partnerships with state governments like New York are key ingredients needed to match the subsidies offered by Asian nations and “bring chip-making back to America,” Mr. Mehrotra said (Source)
The plant will be in Clay, N.Y., about 15 miles north of Syracuse. The timeline is something like: site prep ‘23, construction ‘24, volume production ‘25.
Again, over 20 years to get the biggest number possible, this is estimated to create 9,000 jobs for Micron employees and 40,000+ for third parties in the supply chain.
🇳🇱 Mini-documentary: ‘ASML’s Secret’ (Top 2 most important companies in the world?)
re: the headline, you know that the other one is TSMC, right?
🧪🔬 Liberty Labs 🧬 🔭
The future as predicted from the past… ⚡️
Caption: West Germany's government's forecast for energy mix from 1974 before it was hijacked by the irrational fear of nuclear power
I know, I’m a broken record — get a new pet topic, Liberty! — I’m sorry, it’s just that nuclear strikes multiple chords with me.
It’s technically fascinating, it’s unjustly maligned, it has huge potential to *actually* make the world a much better place (cliché to say the words, but once in a blue moon, it’s true), and seems to be starting to get some traction (it’s harder to be excited about something that seems entirely hopeless).
Back to the chart above: Imagine if that had happened. Not only in Germany, but in lots and lots of other countries. How different would the world be?
The Amazin feat of engineering that is the SR-71’s Pratt & Whitney J58 Engines 🚀
That thing was most efficient at Mach 3.2 (that’s 2455.26 MPH).
Do you know why it’s called the J58?
Because its ‘first run’ was in 1958 (!!! 🤯 — the SR-71’s first flight was in 1964).
This thing was designed with slide rules, and there was probably an abacus involved somewhere 🧮
🎨 🎭 Liberty Studio 👩🎨 🎥
Generative AI + Stable Characters 🧙♂️🧟♀️🧜🏻♀️🧞♂️ 🎨🤖
The possibilities once you can have arbitrary stable characters to play with are endless.
Thanks for shoutout, any day I make it into the news letter is a great day.
I love the German energy mix forecast for a number of reasons, really cool graphic. Isn't it curious though that while oil and gas go to zero, coal generation remains constant through almost the entirety of the forecast period? I would love to know the reasoning.