178: Apple iPhone 13 Event, TSMC!, Horn Chart of Compute, Larry & Sergei, Amazon, Datadog, Omega 3, Semiconductors, Pseudonymous vs Anonymous, and Kinetic Sculptures

"a treadmill that speeds up all the time"

A scholar is just a library’s way of making another library.

—Daniel Dennett

👺Being pseudonymous is very different from being anonymous.

Knowing someone's legal name doesn't necessarily mean that you know them more.

What matters most, IMO, is what someone thinks and shares, your relationship with them, not the specific name attached.

Someone who has been following me under “Liberty” since I chose it as a nom de plume (I doesn’t sound pretentious when I say it because French is my native language…) around 2010 probably knows more about large parts of my life than my parents and friends, because they are not very interested by investing and tech, so I don't talk to them much about these things, yet they’re topics I think about all day long.

Conversely, I’m sure we can think of many example of people writing under their “real” names, but they’re not sharing anything that matters, only reposting links and memes — do you know more about that person?

If you use a pseudonym long enough, reputation — good & bad — accrues to it, and functionally, it becomes a “real” name in most of the ways that matter.

That’s why I don’t feel much temptation to switch to my real name. It’s just a regular name anyway, not as good as “Liberty” 🗽 ¯\_(ツ)_/¯

🗓 I think it’s good to remind ourselves once in a while that the numbers on the calendar have nothing to do with the level of sophistication of technology that we have.

We’re so used to seeing it improve year by year that it seems like an inevitable march of progress, but for very long periods (like tens of thousands of years) during humanity’s history, there was basically no progress.

Even when there was a lot of it, things sometimes took a turn for the worse and we moved backwards.

If the Enlightenment/scientific revolution had happened a few hundred years earlier than it did, maybe the Internet would’ve been created by the 1700s. Where would we be by now?

Or without the step back of the medieval dark ages — I won’t argue about definitions and exact duration, but it lasted a while — we may have terraformed mars and cured the diseases of aging by now. (darn, right?)

🤔 I wonder how big the spike in sales of tarantulas as pets was after the movie ‘Home Alone’ came out…

🥳 Congrats to Patrick O’Shaughnessy on 5 years of ‘Invest like the Best’. What a body of work, what a great showcase of the power of learning in public and non-zero sum value creation for an entire community. When all is said and done, I wouldn’t be surprised if this project had created billions in value — for host, guests, listeners — with a fair share of that captured by Patrick, directly and indirectly. Kudos!

🌏 Yesterday I had a 1.5 hour call with Lillian Li. Incredibly smart and nice.

Electrons and photons going back and forth from Canada to China

My first thought as the call started was that she’s not “over there” if I point in any direction along the horizon, she’s basically under my feet — we’re both standing on the sides of a planet with our heads pointing in different directions, making a kind of ‘Earth sandwich’… And yet the lag in the video call wasn’t that bad at all, I’ve had worse pings in the early days of multiplayer games. Speed of light FTW.

🙊 I often think about when I’ll be really wrong about something in this project.

At some point, I’m sure to do a big angry rant 🤬 about something, and then realize that I was totally wrong about it. Or I’ll share something, only to realize that my brain just went “oops” and I wasn’t supposed to share it. Or maybe I’ll write for a while about a company I admire, and then it’ll come out that they were actually Valeant 2.0 or something of the sort.

On a long enough timeframe, something is bound to happen. I just have to make sure that I’m ready to deal with it graciously, in a transparent manner, and to update my mind rapidly rather than fight reality and get stuck in a defensive posture.

We all gotta train for the big game, because it’s when reputations are made or broken…

🛀 I wonder if wild animals are ever afraid of the dark?

💚 🥃 The price of a couple coffees or one alcoholic drink isn't a bad trade for 12 emails per month (plus 𝕤𝕡𝕖𝕔𝕚𝕒𝕝 𝕖𝕕𝕚𝕥𝕚𝕠𝕟𝕤) full of eclectic ideas and investing/tech analysis.

The entertainment has to be worth something on its own, but for those that care most about the bottom line, there’s optionality:

If you make just one good investment decision per year because of something you learn here (or avoid one bad decision — don’t forget preventing negatives!), it'll pay for multiple years of subscriptions (or multiple lifetimes).

As Bezos would say of Prime, you’d be downright irresponsible not to be a member, it takes 19 seconds (3 secs on mobile with Apple/Google Pay):

💙 Subscribe now 💙

Investing & Business

Apple iPhone 13 Event, My Thoughts

I don’t have much to say about this one. I’m still waiting for the new iMac Pros that run on ARM M1X or M2… And maybe new AirPods. And maybe they’re starting to convince me to get a Watch (if they can ever figure out how to have a blood-glucose monitor in a Watch, it’s an insta-buy for me, but that’s not an easy technical challenge…).

Apple has been doing this a long time, yet people always complain when they only have incremental improvements rather than a big “one more thing” moment that blows everybody’s minds.

This is your reminder that they only have those once in a while (iMac 1998, iPod 2001, iPhone 2007, iPad 2010, Watch 2015), and it’s fine. Scorsese doesn’t make Goodfellas every year ¯\_(ツ)_/¯

At this point, most people upgrade their iPhones every 3-4 years.

The jump from an iPhone 7 or 8 to a 13 is gigantic. Everything will be MUCH faster, photos will look ridiculously better, battery life will be a big improvement, the water-resistance and screen durability are a huge plus, the ultra-wide lens allows really fun photos and videos that you just couldn’t take a few generations ago, the new OLED screens are much nicer, etc.

I’m still looking forward to a big new product category introduction and hope they can pull it off once again, but in the meantime, I’m fine with the incremental improvements to stuff that may not make for sexy demos, but that improves the kind of features that people use all day long (ie. big camera improvement this year), because that matters a lot more than “cool” stuff that almost nobody uses (some companies focus a lot more on that).

✨ Morris Chang ✨ & TSMC

About a week ago, I had the pleasure to do a Zoom call with supporter (💚 🥃) David Rosenthal and Ben Gilbert, co-hosts of the so very excellent Acquired podcast.

Their biggest quality and competitive advantage is that they really really like learning about all this (companies, technologies, histories…) and have that genuine curiosity and ‘smile-in-the-voice’ even when the ‘record’ button ⏯ isn’t pressed.

Their most recent episode is ostensibly about semiconductor contract manufacturing and design company TSMC, but it’s really more of a biography of Morris Chang, the company’s founder and easily one of the most important technologists of the past 100 years:

I loved this line by Chang that they quote:

The semiconductor industry is like a treadmill that speeds up all the time. If you can’t keep up, you fall off.

Implied is that if this happens, you probably can’t get back on… which is what Intel is trying to do. They have a chance to pull it off, but only because they’re so big and have so much expertise and access to capital and are a national champion. But even with all this, it’s not a given that Gelsinger can pull it off.

If they were a less successful and important firm, they’d have no chance.

Gavin Baker on the Semiconductors Situation

Interesting interview with Gavin Baker, here’s my highlights:

Everybody thinks this time it’s different, but it’s never different. As lead times are increasing, I promise customer buffer inventories are increasing. If you’re a giant internet company, to provide services you need a lot of CPUs, GPUs and memory. You don’t want to run out of those processors and have the company’s growth curtailed. So you’re building inventory everywhere. If you’re a big handset company, you’re building inventory. Everybody is building inventory, and there is also something called double ordering. That means if you’re a handset company, you’re placing an order with more than one semiconductor company for what you need because you’re not sure you are going to get it. The semiconductor companies are always thinking it’s not happening. But it’s happening.

Semiconductor companies are straining to meet this demand that’s actually growing stronger and faster than true end demand. And then something happens. Who knows what it is? Maybe it’s the crypto currency market rolling over. Maybe it’s a big smartphone company coming out with a device that’s not as cool as it usually is. Maybe it’s the Delta variant slowing down economic growth. But all of a sudden, lead times go down one tick, say from twenty weeks to nineteen weeks. So of course, customers reduce their inventories. All of a sudden, demand flips from above natural to below natural, and then it spirals the same way down as it went up. That’s why you have these wild inventory cycle driven swings in end demand. It’s pretty clear that if we have not already seen the peak of this inventory cycle, we’re very close. But once we’ve gone through that, the open question is if we’re in a true capacity cycle. [...]

Semiconductors are cyclical and economically sensitive, but long-term they grow faster than GDP. Now, artificial intelligence is going to raise the semiconductor intensity of global GDP. In other words, if in the past the semiconductor industry grew at 1.5x global GDP, now it’s going to grow 2x to 3x because AI, by its very nature, consumes so much more semiconductor content than software written by humans. Therefore, it may be possible that the world is structurally short of capacity, and we’re in a capacity cycle, driven by the confluence of AI and all these other trends: the electrification of cars, autonomous vehicles, crypto mining. All these mega trends will be driving semiconductor demand, but they’re all dwarfed by AI. Yet, we’re only going to be able to make that judgement once we’re on the other side of the current inventory cycle.

What’s interesting with AI’s unquenchable thirst for compute is that it’s going up faster than the rate at which chips are improving (model size is doubling every few months, and more and more organizations are adopting AI/ML processes, so there’s this double-whammy).

🤘 The ‘Horns’ Chart of Compute 🤘

I really like this chart by the fine folks at NextPlatform. Here is context:

one of our favorite charts that we have ever built, which shows the relative price/performance of server capacity by quarter (adjusted for inflation, on the left axis) plotted against the aggregate amount of relative performance in the total number of servers shipped each quarter (on the right axis). Nice exponential curves showing the elasticity of consumption since 1999

It takes a moment to get it, but once you do, it’s beautiful 🤩

Larry & Sergei

Sometimes I wonder what they’re up to. That’s all1.

Amazon Hiring Everybody, Redux

In edition #173 I wrote about Amazon’s strategy of basically hiring everyone. Well, looks like they still have some people to hire:

[Amazon] announced that it is providing an additional 125,000 local employment opportunities throughout the U.S., on top of the 40,000 corporate and technology jobs announced earlier this month. The roles in fulfillment and transportation offer an average starting wage of more than $18 per hour—and up to $22.50 per hour in some locations. [...]

In 2021, Amazon opened over 250 new fulfillment centers, sortation centers, regional air hubs, and delivery stations in the U.S., and will open over 100 more buildings in September alone… Since the beginning of the pandemic, Amazon has hired more than 450,000 people in the U.S (Source)

$18/hour is going to put pressure on competitors…

Datadog (💾 🐶) Primer with Peter Offringa

Friend-of-the-show Peter Offringa (💽) did an episode of Business Breakdowns on Datadog, and did a great job of succinctly (40 minutes without the intro) describing where the company comes from, what it does, who else operates in the space, and where it is going.

If you have an interest in the dog and his data, check it out:

💙 Subscribe now 💙

Science & Technology

‘The four most important charts in science & technology’

Click through to the Tweet to see larger versions.

It’s log graphs of 1) number of transistors on chips, 2) price of solar modules, 3) the cost of sequencing a human genome, and 4) the price of lithium-ion batteries.

Here are the same charts, but on linear scales rather than log, just for the lolz.

Someday, your ring may tell you when you’re pregnant

Sleep-tracking company Oura did a study with the University of California San Diego to determine is their wearable could detect pregnancies in the biometric data that they collect. Interesting results:

Oura’s nighttime temperature data was able to help identify pregnancies early, on average 9 days before these individuals received a positive at-home pregnancy test and only 5.5 days after conception (self-reported). [...]

All 30 pregnancies showed a clear temperature increase following self-reported conception that was identifiably higher than the increase that typically occurred during the luteal phase of their menstrual cycle.

I quite enjoy my Oura.

It helped teach me what had the biggest impact on my sleep quality, and the tight feedback loop helped change some habits.

The temperature-tracking has also been an early indicator when I’ve been fighting off a virus, sometimes showing an anomaly a day before I felt ‘off’.

Omega-3 101, 102, and 103 (and maybe 201 too)

Great podcast interview of Dr. Bill Harris by Rhonda Patrick. I’ve learned a bunch listening to this, and I’ve been educating myself on Omega 3s (EPA and DHA, ALA not so much) for close to 20 years…

I highly recommend it, and even if you don’t need that level of detail, I highly recommend that you eat plenty of fatty fish (salmon, sardines) and take multi-grams of supplements daily. It’s one of these low-hanging fruits for which the risk/reward is incredibly asymmetric. Here’s the pod:

💙 Subscribe now 💙

The Arts & History

Kinetic Sculptures

I recently saw a video of a really cool wind-powered kinetic sculpture made by Anthony Howe (above).

Peter shared this great interview with David C. Roy, a very skilled kinetic sculptor who not only shows many of his pieces, but explains what goes into designing and building them. Great stuff:

If you want to learn more, or even buy some of David’s art, this is his official website:


I mean, they do have voting control over an almost $2-trillion market cap company that is deeply embedded into almost everybody’s life. They’re powerful people on Earth at the moment…