310: Transdigm, Roper, Samsung in Texas, TSMC, James Webb Space Telescope, Google DeepMind, F-22 and SU-57, and Christopher Nolan's Oppenheimer
"we emerge from the summer breeze"
Creativity is contagious — pass it on!
🌳🍼👶 I was thinking about the famous Richard Feynman video where he explains that trees come mostly from the air, not the ground…
This time, I thought a few steps further in a way that I hadn’t before: if carbon-based life-forms are either plants that grow from the air, or animals that either eat those plants or eat the animals that eat those plants, and we humans are one of these animals…
It means that most of the carbon in our bodies comes from the air!
If you trace it all the way back, we emerge from the summer breeze.
Poetic, ain’t it?
🛀 🛰🔭 I was thinking about the James Webb Space Telescope and how special it is to have a piece equipment of such sophistication floating 1.5 million kilometers from Earth.
When writing about it, journalists will often talk about the dollar cost of it.
“JWST cost $10 billion dollars! Wow!”
That’s not the best way to impress me with it. It kind of implies that anyone with $10 billion could go out and buy one.
What’s most impressive to me is that almost everything in it must be bespoke, custom, and almost unprecedented.
When you build a car or a fighter jet, you get to make many copies and go up the learning curve, and down the cost curve.
If Lockheed had only made a single F-22 Raptor, the price tag would be like $40 billion. But if you make 1,000 or 10,000, you get to spread the heavy upfront fixed costs over the units (there won’t be that many F-22s — there’s only about 200 so far — I’m just illustrating things).
The James Webb doesn’t get that, and the first unit REALLY needs to work reliably for decades (hopefully) because it’s going to be parked in the vacuum of space 3x+ farther away than the moon!
So not only do you have to write a bunch of code that needs to be super reliable, but that’s not the hardest part because you can usually update software unless you’re really unlucky and all your failsafe doodads don’t work and you turn the thing totally inert.
But the hardware! The advanced precision manufacturing that went into every single part of the thing! How much effort must’ve gone into spinning up processes to make single-digit units (I assume they had some spares and prototypes) of a bunch of parts that nobody had made before and nobody really has experience on? Everything needs to fit together perfectly at the end of a multi-decade process..? And then this needs to be folded up, put on top of a rocket, and blasted to space on a wild, shaky, high-G ride! 🚀
That’s what impresses me about it, not the dollar amounts.
💚 🥃 There are now over 60 people in the Liberty Labs 🧪🔬🧬 🔭 Discord and it hasn’t even been a week! 🙏
Liberty’s Highlights is reader-supported. To support my work, consider becoming a paid subscriber. This also gives you access to the Liberty Labs 🧪🔬🧬 🔭 Discord.
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 💳 💴
👶🏻 How much can be accomplished in one lifetime 👴🏻
Last week, the news of Dee W. Hock’s death — he founded Visa — made me reflect on this line by Bill Gates that I’m sure you’ve heard:
Most people overestimate what they can do in one year and underestimate what they can do in ten years.
Ten years. How about a lifetime?
To think that credit cards went from “Hmm, I think this could be a good idea” to ubiquitous as the platform on which trillions of dollars in payment flows and market cap are built (speaking of which, I did a podcast about the payment industry with MBI).
Someone born before the Wright brothers could watch the moon landing on their TV, someone born when Asimov was writing science fiction about computers could walk around with a touch-screen super-computer that is always connected to a global network of billions of computers via radio-waves in their pocket (oh, and the thing is also a much better camera and walkman and telephone and ETC ETC ETC than anything that was available not long ago).
Things move fast! What will it be like in 20 years? 40 years? ¯\_(ツ)_/¯
Interview: Rob Small of Berkshire Partners on Transdigm 🛩
Rob Small joined the Transdigm board in 2010. Before that he did both private and public investments in the company.
It’s an interesting overview of the operational and financial model, and helps explain why — unless another bidder overpays — TDG should be able to win most auctions that they get into at prices that result in impressive IRRs. TDG 0.00
🏭 Roper’s post-industrial composition 💾
In edition #288 I wrote about how Roper was selling 51% of its industrial/cyclical businesses to private equity for $2.6bn.
They recently released their earnings for Q2 and gave some updates on the new composition of the business after this:
At the top are the 3 new segments, and below you can see how the revenue breaks down between software and non-software, and how much is recurring and reoccuring (which isn’t quite the same but implies relatively predictable after-market sales on a large installed based).
Quite a different mix from 10 or 20 years ago… Net working capital is now -17% thanks to upfront payments for subscription software. ROP 0.00
⭐️ Samsung considering $200bn (!!!) expansion of 11 chip plants in Texas 💰💰💰💰💰💰💰💰
The South Korean company, the leading maker of memory chips, laid out potential plans to spend almost $200 billion on 11 plants in a series of filings in the state. Two of the units would be in Austin and nine would be in Taylor, Texas, where Samsung has already unveiled plans to spend $17 billion on an advanced facility.
Nothing’s certain yet, though, they’re just considering it.
“We currently do not have specific plans to build at this time, however, the Chapter 313 applications to the State of Texas are part of a long-term planning process of Samsung to evaluate the viability of potentially building additional fabrication plants in the United States,” Samsung said in a statement. (Source)
Dan Nysdetd writes: “Taiwan media lament that Samsung's planned US$200 billion investment in 11 chip fabs in the U.S. could threaten TSMC's business due to its the sheer size of the operations. Last year, 65% of TSMC's revenue was from North America.” TSM 0.00
Interview: Dan McMurtrie 🐜
Another really good one by Dan. He’s a very interesting thinker, goes to the heart of things and is able to map out pretty clearly what the problems and potential solutions are, while staying honest about what’s impossible to know.
Another great interview by friend-of-the-show and supporter (💚 🥃) Jim O’Shaughnessy!
🇺🇦 TIL that Ukraine’s GDP per capita is around $5,000 USD/year (well, before the war 😥)
I’m sharing it because it was a surprise to me. I’m not sure what I would have answered if someone had asked me to guess without looking it up, but probably higher than that.
(Of course, this is in the rearview mirror and doesn’t reflect the current reality)
🧪🔬 Liberty Labs 🧬 🔭
Interview: Demis Hassabis, DeepMind CEO & co-founder 🤖
Really interesting interview, particularly when it comes to the potential applications of DeepMind’s AIs for large scientific problems like protein folding (AlphaFold 2 has largely cracked that one) and controlling plasma inside of nuclear fusion reactors, among others.
I’m both very excited and worried about the development of ever more powerful AIs (on the way to AGI), and the fact that it’s all happening faster than even experts expected not long ago makes it even more worth paying close attention to. GOOG 0.00
Germany dithers on its last 3 nuclear power plants 🇩🇪☢️
After making all kinds of claims that it was “technically impossible” to keep the last 3 nuclear power plants in the country running, and thanks to the support of the Green Party who de facto prefers coal & gas & Putin to uranium, it looks like there’s now a slim chance that reason may prevail:
Economy Minister Robert Habeck has commissioned a new stress test for Germany’s energy infrastructure, after an earlier assessment concluded that the country’s supply wouldn’t be endangered next winter and that nuclear power plants wouldn’t be needed.
Seems absurd to me that “they won’t be needed” is an argument as the country is heavily dependent on foreign gas and is restarting shut down coal plants as fireplaces are selling out… Why should the goal be to barely scrape by with fingers crossed rather than have a solid surplus of clean power?
Behind the scenes, the coalition parties in Scholz’s government have indicated they may be willing to agree to a political deal that would allow passage of the controversial measure.
Former F-15 and F-16 pilot shares his thoughts on F-22 and SU-57 fighter planes
If you want a more detailed comparison, this video goes into a lot more depth (with a puppet!).
🎨 🎭 Liberty Studio 👩🎨 🎥
Some details on Christopher Nolan’s film about J. Robert Oppenheimer and the Manhattan Project ⚛️💣
I soooooo hope this one will be good. A major blockbuster about physicists and engineers! Maybe it’ll inspire some young people to go into STEM?
Richard Feynman is one of my heroes (I know a lot of people say that, but I won’t be a hipster about it and not like what I like just because others also like it…) and he’s part of this story. I don’t know yet which actor will play him, but when you look at the cast list, there are so many great actors that it’s very promising:
Cillian Murphy as J. Robert Oppenheimer
Emily Blunt as Katherine "Kitty" Oppenheimer
Matt Damon as Leslie Groves
Robert Downey Jr. as Lewis Strauss
Florence Pugh as Jean Tatlock
Rami Malek, Benny Safdie, Josh Hartnett, Dane DeHaan, Jack Quaid, Matthew Modine, Dylan Arnold, Olli Haaskivi, Alden Ehrenreich, David Krumholtz, Michael Angarano, Kenneth Branagh, David Dastmalchian, Jason Clarke, Louise Lombard, Scott Grimes, Christopher Denham, James D'Arcy, David Rysdahl, Guy Burnet, Danny Deferrari, Josh Peck, Harrison Gilbertson, Emma Dumont, Matthias Schweighöfer, Gustaf Skarsgård, Devon Bostick, Alex Wolff, Tony Goldwyn, Trond Fausa Aurvåg, Gary Oldman, Josh Zuckerman, Olivia Thirlby, Casey Affleck
I mean, they just throw Gary Oldman and Casey Affleck at the end…
I’ve read the Kai Bird bio of Oppenheimer on which the film is based and really liked it. I wonder how they’ll do the 🍎 scene 🤔
I love that Matt Damon will play General Leslie Groves, because Groves was incredibly important to the success of this project and deserves the spotlight (there’s a book about the partnership with Oppenheimer called ‘The General and the Genius: Groves and Oppenheimer’ by James Kunetka — I haven’t read it, but friend-of-the-show and supporter David Senra (💚 🥃) did a great podcast about it).
Another great book on this: ‘The Making of the Atomic Bomb’ by Richard Rhodes (Rhodes also wrote a follow-up about the development of the hydrogen fusion bomb…).
If you’ve never seen the classic bit of interview with Oppenheimer, it’s a must-see:
Your thoughts [Liberty at the top] on the "JWST cost $10 billion dollars! Wow!" and how the dollar cost misses the point of the project.
-- 👏-👏-👏-👏 That is me slow clapping for your thoughts and sentiment. The dollars misses the point, or is the wrong metric to judge the project by. The dollars are just dollars, but what they accomplished 1.5M KM away over multiple decades is spectacular/amazing/breathtaking! ⚡😲🛰
Seems a bit bonkers from Samsung to put all their eggs in one basket/state line that. Even one country is a little extreme. Staffing it is going to be interesting as well given Texas is not exactly everyone's cup of tea. Will be interesting to see how that plays out.