46: My 1st Interview, Alibaba's Cluster of Billionaires, Google Maps, Semiconductor Manufacturing 101, Opioids, Color Film, and Mad Men
"shooting 4k+ resolution on an analog medium"
“Smart people learn from their mistakes. But the real sharp ones learn from the mistakes of others.”― Brandon Mull, Fablehaven
I have this printed out and taped to my bedroom's mirror as a reminder to my wife & I.
→ If you’re a US citizen, please vote. I mean it. ←
➽ Something to add to your probability tree (I don’t know what will happen, but I think this is a real possibility):
If T. loses the election, he won't even have to pretend to care about the country until January (during the lame-duck session), and will probably actively try to make many things as bad as possible — or at least not make them better — so B. has a worse hand to play. Don't expect stimulus or rational action on pandemic or much else positive... He'll try to poison the well.
“He wouldn’t. Doesn’t his big ego care about his legacy?”
Like how nothing in biology makes any sense unless seen through the lens of evolution by natural selection, nothing about Trump makes any sense unless seem through the lens of malignant narcissism and sociopathy.
So while the moment-to-moment actions of such a person can be very unpredictable, the high-level motivations are extremely stable and uni-dimensional. A big blow to his ego will be met with vengeance and lashing out.
He does care about what is said about his legacy, and know he’ll be able to say that everything he did was great and blame any problems on others, as he has done forever, and that millions of his fans will go along with it. Since the market is his favorite barometer of success, he would probably love nothing more than to see it crash and say it’s because of B’s election.
There’s only ego and attention (aka narcissistic supply).
In any case, I look forward to the day when I can go back to ignoring politics and thinking about more interesting things. I remember how before this, I could spend weeks or maybe even months without hearing or thinking about the president of the U.S.…
Investing & Business
My 1st Interview (TXN, NVDA, CRWD)
I’ve never done this before. But Rob from Koyfin is a really cool guy, and I figured, why not give it a try?
Behind the scenes: So Rob asked me what I wanted to talk about, and I decided that Texas Instruments and Nvidia may be good topics, because I have been learning about them, and they have interesting tech and business dynamics.
So I wrote down some notes about them. Kind of an outline, just so I’d have something to look at if nerves or whatever made me forget what I wanted to say.
I work alone, so I don’t use video conferencing much. After we started the interview over Zoom, Rob shared his screen with me, so that I could see what he’s doing on Koyfin.com as we spoke.
When he did that, his screen share went full screen on my screen, hiding my notes.
For a split second, I wanted to hit ESC to minimize it, or considered maybe mousing around to find somewhere in the interface to minimize his screen-sharing to a window, but I was speaking at the same time and was in danger of losing my train of thought if I tried to do that at the same time, and knew that if it didn’t work (if ESC killed the screen share, f.ex.) it’d be a problem, so I just kept going.
I did the whole thing without my notes.
Then after we finished talking about these two companies, Rob asked “anything else you want to talk about?”, and the smart answer probably would’ve been “no, I think we’ve covered a lot”, but I was having fun, and spur of the moment said, “well, maybe Crowdstrike” because I had been learning a bit about it recently.
I did that segment without notes or any preparation.
So if it’s all a jumbled mess, that’s why. I know I just gotta let it go, that’s what the medium is, you speak into a microphone and every word is a bell you can’t unring, unlike with writing where you can go back, edit, clarify, look things up and come back, etc.
Oh well, I did my best. I hope some of you enjoy it.
Alibaba’s Cluster of Billionaires
Interesting thread by Dennis Hong of ShawSpring Partners:
One of my friends had a great insight: "I think one aspect of Alibaba's success that has been overlooked is that Jack Ma spread the equity generously amongst the leadership team. It's probably the only internet company that's produced 20-30 billionaires instead of just 2 or 3."
"AMZN which is worth twice the market cap of BABA has only produced two billionaires: Bezos and his ex-wife. If you have 20-30 billionaires that form a small elite and control the company, they will never need to leave and you can actually transition leadership really well."
Bloomberg: Alibaba’s Co-Founder Lucy Peng to Get $5 Billion Richer on Ant IPO
"I think ultimately the best founders understand it's okay to give up a lot of your equity to keep your best people forever. The Huawei founder diluted himself from 100% to 1% to build the partnership. That just goes so much against human nature."
“There is this old Chinese saying - 财聚人散 财散人聚 which roughly translates as: If you keep all the money together, the people will scatter. But if you scatter the money, the people will come together.”
Startups: Boy Bands vs Punk Bands
Did You Know the Concept of a “Compounder” is Proprietary Information?
This story is just bonkers. A partner at a multi-billion dollar hedge fund threatened to sue a Substack writer because they claimed they were using “proprietary information”. What kind? Did they steal some HFT algorithm or a database?
Not exactly… They were writing about companies known as “compounders”. Mario Gabriele tells the story:
When I asked what precisely she objected to, she struggled to articulate particulars. Later, she claimed that she and her co-professor had created the definition we used in the article that compounders grow "20% or more over ten years." She also argued that the patterns we identified at the bottom of the piece ("good to great," "cluster horses," and "reload for growth") were proprietary. [...]
When I followed up the next day, Ana accused the article of being "largely plagiarized," and she demanded the "project" stop unless I wanted the matter escalated to Durable's general counsel or Columbia Business School. [...]
I was frustrated in the extreme. I knew that we had not done anything wrong, that “compounders” is a well-known business concept, and that we had not plagiarized or improperly taken “proprietary” information. But Ana had twice threatened to take the matter to Durable’s General Counsel. Durable is a six billion dollar hedge fund. The Generalist is a start-up newsletter with $0 in revenue. Six-billion dollars can stare you down, no matter how right you are. It can cow you and shame you and make you slink to your bedroom in the middle of the afternoon and cover your head with a pillow and think about crying, which I did.
To me it seems like a good example of this very undesirable human trait: someone (the prof) making a bad decision in the moment, and then doubling down and doubling down again and again, instead of admitting they're clearly wrong and apologizing.
It also seems to me like a certain professor didn’t know about the Streisand Effect.
You can go read the whole saga at The Generalist (feel free to subscribe to support David against Goliath), and read the article that caused all this controversy.
Acquired: Google Maps (Podcast)
I enjoyed this episode of the Acquired Podcast (I’ve been slowly making my way through their archives) about the acquisitions that lead to the creation of Google Maps.
Ben and David cover the series of three 2004 Google acquisitions that formed the core of the Google Maps we know and love today: Where 2 Technologies, Keyhole and ZipDash. From nearly zero adoption between the three companies at the time of acquisition to well over 1 billion users today
One thing that I always think about when the topic of many of the all-time-great acquisitions pops up is “how much was actually acquired and how much was built by the parent company after the acquisition”.
It’s a hard question, because a lot of the time, even if not much is acquired and almost everything is built afterwards, it may not have been built without the initial spark/prime mover that was the acquisition.
But on the other hand, I think a lot of the time, too much value is assigned to the original acquisition, as if the same outcome was inevitable even without the deal. You hear about how they paid X for something and now it’s worth 100X, but it matters to the return if after the acquisition they invested 20X to get it there.
Would Instagram be Instagram if they had stayed independent? Would Booking be as big? Would Youtube or Android have won dominant positions as stand-alones? How about National Indemnity for a more clear-cut case… Probably ‘yes’ for some of them, but likely ‘no’ for others. It’s hard to untangle.
360 or 365?
Science & Technology
Color Film is Amazing Technology
When you think about it, color film is a pretty amazing technology. Whenever the quality of film stock peaked, 50 years ago or whatever*, they were basically shooting 4k+ resolution on an analog medium that required a bunch of chemicals to process...
We don’t think about it anymore, and when we do take it totally for granted, but it’s a pretty great achievement of human ingenuity.
It’s always hard to compare analog to digital directly, but I’ve seen various napkin calculations that put 35mm film somewhere above 4k resolution. Here’s one estimate (of course it always depends on the film stock and sensitivity/grain size):
A "35mm" frame is 36x24mm in size. Look at the resolution spec for some films and lenses. Some films were rated at nearly 200 lines/mm, but some much less. There was a tradeoff between sensitivity and grain size. That added noise and lowered spacial resolution of more sensitive films. Lenses also cover a range. Let's say roughly 50 lines/mm would be "good", and 100 lines/mm astonishingly superb. Of course that's only at the optimum f-stop and camera mounted and held very still.
You said top equipment, so let's see what 75 lines/mm comes out to as a starting point. A "line" is actually one complete light-dark cycle, so you have to allow for at least 2 pixels per line width. So the 75 lines/mm becomes 150 pixels/mm, which means a full 35mm frame would have 5400 x 3600 pixels = 19.4 Mpix.
For comparison, 4k is 3840 × 2160 for TVs and 4096 × 2160 for movie projectors.
If you want to go deeper this rabbit hole, there’s more about the various pros and cons of each medium here.
*Some seem to argue the peak was before the 1960s.
‘No other country in the world has seen a surge in opioid overdose deaths like the US.’
Via Max Rosner, on Why is life expectancy in the US lower than in other rich countries?
Great Presentation by a Semiconductor Engineer on ‘Manufacturing Chips’
There’s two versions of this by the same presenter floating around, but I recommend the 2013 version if only because it’s a bit more recent (I think the other was from 2008):
For those of you curious about some of what’s involved in manufacturing these magical pieces of silicon that run so much of the modern world, without which you wouldn’t be reading these words right now, I think it’s a good starting point. Youtube is full of treasures like that…
TIL About Burls
During a walk in the woods, I saw this thing. I tweeted: “I wonder what that is. Some kind of tree cancer?” and got an education about burls:
A burl results from a tree undergoing some form of stress. It may be caused by an injury, virus or fungus. Most burls grow beneath the ground, attached to the roots as a type of malignancy that is generally not discovered until the tree dies or falls over. [...]
Burls yield a very peculiar and highly figured wood, prized for its beauty and rarity. It is sought after by furniture makers, artists, and wood sculptors.
@roboco23 posted a piece about how rangers have to stop thieves from stealing burls in federal forests:
Burls contribute to a forest’s complex ecosystem of growth and regeneration. In the eyes of collectors, their rich and intricate grains make them prized pieces that can be crafted into tables, bowls, and other objects. Taking any plant from a national park is a violation of multiple U.S. laws; taking and selling old-growth redwood, which takes hundreds of years to develop, strikes Pero as particularly egregious.
Tren Griffin even shared his OG burl butcher block.
The Arts & History
I like how this shot of a random field I walked by came out. Shot on iPhone 12 Pro, adjusted the colors a little bit, but not much. Mostly untouched.
Note that Twitter and probably Substack recompress the original, and probably don’t support the HDR and wide color gamut that the iPhone 12 Pro does.
I can’t wait until all the big sites support fully these features. Imagine how much bandwidth and storage would be saved if all JPEGs online were replaced by the more modern HEIF format that takes half the space at equal quality and has more flexibility/expandability (HEIC is the container for HEIF and uses the key frame compression tech from h.265/HEVC, which is one of the top current video compression CODECs — if you’ve ever watched 4K on Netflix, it was using HEVC).
Rewatching Mad Men
Over the past couple weeks I've started a rewatch of 'Mad Men' (2007, AMC, streaming on Prime) with my wife.
It's her first watch, and it’s kind of nice for me to see it through her eyes.
I’ve always considered it one of my favorite shows, but seeing it again is reminding me why. Like Deadwood, almost every scene is about something other than what the characters are saying, and it’s firing on all cylinders (writing, acting, set design, cinematography, music, costumes, symbolism & themes & metaphors…).