110: Shiny New Object Syndrome, Constellation, Twilio, Netflix, TSMC^2, ARM V9 & Nvidia, Payments, Visa & Crypto, Cognex, Intel's Nanometers, Jupiter Moons & Tiger Stripes
"What’s the cure, what’s the cause?"
Try to accept hard truths a little faster than other people do.
It is a great advantage.
🗣 If you’re trying to do something — like a big ambitious project or have a career in X or become skilled at Y or whatever — it’s often a good idea to talk to someone who has done what you’re trying to do, or something similar.
But don’t ask for advice.
Most people aren’t that good at giving advice, or they’ll default to a few clichés you’ve already heard. Not because they have bad intentions or are not really trying, but because we’re not very good at distilling our life’s story into actionable highlights on-the-fly.
Instead, just ask them about their story, what they did to get there and what their days are like now.
Do the work of extracting the advice from this yourself, I think it’ll work out better that way.
✍️ There’s a feeling I get when I’m creating anything, I wish I had a word to describe it.
Somewhere in there, there’s this magical moment where I realize that it’s moving from a pile of crap that isn’t really anything to something that is taking form, something that wasn’t there before and that just may be decent.
That moment feels really good.
I’ve had it in everything that I’ve dabbled in: Making maps for Doom and Quake as a kid, cobbling together some riffs on the electric guitar while my friend played drums, and writing these newsletters.
To me it’s the best part of creation.
Coming up with ideas is neat, but you don’t know yet if it’s going to work. The finished thing feels good to look back on sometimes, but not nearly as good as that “in between” stage where you see the statue start to emerge from the marble block for the first time.
I don’t know if it works the same for others, but if you also like that feeling and haven’t had it in a while, I want to encourage you to start new projects.
Find something where you are at the bottom of the learning curve and start climbing. It doesn’t have to be a huge thing, but just hone some skills, create something, apply some craftsmanship. I don’t think you’ll regret it.
🤔 I’m surprised the baby monitor industry hasn’t been disrupted by new entrants that takes more advantage of the smartphone supply chain for better, cheaper parts and better UI and connectivity.
How much better would these be by simply using phone screens for the monitor and phone camera sensors to capture the video, rather than whatever crap they’re using? And how about an interface that is a notch above a 1992 VCR 320p screen overlay?
Maybe it has happened in the past 3 years since we bought our baby monitor — maybe we missed the revolution — but what we have is pretty terrible and was expensive, and it was the top model recommended by Wirecutter. ¯\_(ツ)_/¯
😔 So many kids won’t know their grand-parents because of 2020.
That makes me really sad.
Investing & Business
Shiny New Object Syndrome
This is really common. One of my fave investors, someone who I consider a mentor and who I learned a lot from, seemed to suffer from this.
He was great at finding the right businesses and buying them well, but then a year or two later when I congratulated him on how well the stock did, it turned out that he had sold a while ago to buy something else instead.
I think he was just so good at scouring the Earth for ideas, and doing intense research, that he had a high supply of new shiny things that he just couldn’t not own.
But to make space, something had to go, and this probably lead to higher churn than was optimal (I don’t think the new things necessarily were always better than the old ones).
What’s the cure, what’s the cause?
That’s tough. There must be some anchoring going on, in that if existing positions are working well, it can be hard to buy more at higher prices even if rationally you can show through analysis that they’re not necessarily more expensive than at a lower price.
There’s also the attraction of novelty. It’s more fun to learn about something new than do maintenance on something you’ve owned for years.
I think my way of dealing with this is to take relatively small new positions. I try to plant seeds, and if they work out, they’ll grow into bigger, full positions over time, and if they don’t, well, at least they were fairly small.
My biggest positions are things I’ve owned for a while (the concentration is the outcome, not the input, as Elliot Turner would say) and I scratch my novelty itch with this “many smaller bets” side of my portfolio.
Netflix Password-Sharing
Bazinet said that password sharing costs U.S. streaming companies $25 billion annually in lost revenue, and Netflix owns about 25% of that loss. [...]
Netflix began earlier this month testing a new crackdown, in which it is forcing users of its service outside the account holders IP address to use third-party authentication. (Source)
I haven’t seen the napkin used to do the math on this, but it sounds plausible and in the ballpark.
As I wrote in edition #108 on the topic of Autodesk’s software piracy:
it sucks oxygen away from competitors. It’s harder for them to sell their product when they’re competing with both the paid and a free pirated version of the dominant OS/app.
That oxygen-sucking one is prob similar to how Netflix thinks about password-sharing. They may not be paying us, but at least they are watching us instead of a competitor, and forming the habit, which increases the chances that they’ll pay later (either because they get tired of being on someone else’s account or because Netflix tightens the screws on the practice).
But there’s a time to sow, and there’s a time to reap…
h/t Jerry Capital (don’t DM him and ask for his Netflix password)
Constellation Software Annual Meeting
I expected to lose one AGM to this pandemic, not two… This sucks, but hopefully the following year it can be held in person and I can meet a bunch of you guys & gals there!
The Constellation Software annual general meeting of shareholders will be held on Thursday May 6, 2021 at 8:00 a.m. using a virtual meeting format, with proceedings conducted solely via live audio webcast. (Source)
Twilio’s Value
Good line by Byrne Hobart:
Twilio's enterprise value approximates the net present value of the cost of all the engineer-hours it would take to rebuild the relevant subset of their features in-house for every single user.
That’s a good way to think about a whole crop of API/infrastructure companies.
Don’t Feed the Value Investor
My inner value investor was pleased recently.
I bought a new skillet on a 73%-off sale ($30 CAD instead of 190) at Canadian Tire (the only time I go is for their huge loss-leader sales), and one of those outdoor chests/boxes to store deck and pool stuff, on a 25%-off sale ($150 CAD instead of 200).
Then I sold an IKEA desk on Kijiji (online classifieds like Craigslist in Canada) — it’s a Fredrik model where you can adjust the height so that it works as a sitting or standing desk — for the same price I had paid for it almost 10 years ago (free desk!).
That’s the nice thing about buying some durable things second-hand right at the bottom of their depreciation curve. I expect the same to happen with some of the large, durable toys we got for the kids (a big plastic toy-kitchen)… They lose 2/3 of their value quickly, so if you buy it used for that 1/3, you can probably resell it for about what you paid. Free toys!
While I’m at it, here’s a trick to save a ton of $ on razorblades:
It’s not the most professional video — I hope the guy is wearing something below the frame — but the trick works.
Years ago I started doing it and ended up using the same cartridge for like a year, just by stropping the blades on my arm.
It’s basically just a way to do what you’d do with a straight razor and a leather belt — try to straighten and polish the geometry of the very fine edge of the blade, a part so thin that it’s actually kind of soft, so it’s easy to ‘guide’ back into shape — but with modern blade cartridges and your arm (made of leather, when you think about it).
But don’t tell anyone, the razorblade industry would collapse if enough people did this…
TSMC Increasing Wafer Prices 25% “Seasonally”
If I understood the fairly garbled and broken-English Google Translate version of this news piece correctly, TSMC is planning to increase its wafer prices by $400 USD for some customers, or 25%.
Sounds like this will be a temporary increase, most likely only until the shortage situation improves, but this is going to make some waves in the semi world.
Deep Dive (260 pages) on the Payment Industry
It is what it is:
Payments, Processors, & FinTech: If Software Is Eating the World… Payments Is Taking a Bite (Credit Suisse)
h/t Nomad Capital
Risk/Compliance
The older and more cynical I get the more I think that risk / compliance functions at a lot of banks are basically "If you can demonstrate the appropriate social class and industry experience then well YOLO nobody actually reads or maths what you tell us."
Visa to Allow Transactions in Crypto USD Coin (USDC)
isa Inc said on Monday it will allow the use of the cryptocurrency USD Coin to settle transactions on its payment network [...]
The USD Coin (USDC) is a stablecoin cryptocurrency whose value is pegged directly to the U.S. dollar. (Source)
It makes perfect sense. It doesn’t cost Visa much, and they should accept almost everything on their rails.
This isn’t really a sign that they’re backing X or Y, but rather that all the experimenting will happen “out there” and that whoever wins, Visa can support it and plug it into their network.
I expect them to keep adding some of the top crypto-coins over time.
Nobody has more fun on annual reports than Cognex
The CEO and CFO dress up as characters and do elaborate photo shoots parodying various media franchises.
Science & Technology
‘Stretch can shift up to 800 boxes an hour, comparable to a human’
A new robot prototype by Boston Dynamics, which seems less optimized for cool dances and being kicked around, and more for business:
Boston Dynamics claims Stretch can move up to 800 cases an hour, a through-put rate that’s comparable to that of a human employee. High-capacity batteries mean Stretch can operate for eight hours at a time before it needs recharging. [...]
The company says the robot can be operated by anyone with just a few hours of training, and that its mobile base means it can slot into spaces designed for humans. Will it work? We’ll only find out once Stretch gets to grips with the job. Boston Dynamics says it’s currently looking for customers to pilot test Stretch, and is aiming for commercial deployment in 2022. (Source)
We often hear about how self-driving trucks will disrupt the very very large industry of Moving Stuff Around™, but in between all that driving are warehouses, another huge part of the economy, and automation is coming there too.
It’s always the edge cases that make robotics hard, but there’s little doubt that it’s more a question of “when” than of “if”.
ARM V9: Security & AI
ARM (I still refuse to write it Arm, it just looks stupid that way) is unveiling its new V9 architecture, or Armv9 as they write it (they’re bad at picking how to write things). The best context comes from this Next Platform post, IMO:
as we watched the Vision Day presentations by Arm chief executive officer Simon Segars and the rest of the Arm tech team, they kept talking about pulling more vector math, matrix math, and digital signal processing onto the forthcoming Arm V9 architecture. And suddenly, it all finally became clear: Nvidia and Arm both believe that in a modern, massively distributed world all kinds of compute are going to be tailored to run analytics, machine learning, and other kinds of data manipulation and transaction processing or preprocessing as locally as possible and a single, compatible substrate is going to be the best answer to creating this malleable compute fabric for a lot of workloads. What this necessarily means is that both companies absolutely believe that in many cases, the applicability of a hybrid CPU-GPU compute model will not and cannot work.
In other words, Nvidia’s GPU compute business has a limit to its expansion, and perhaps it is a lot lower than many of us have been thinking. The pendulum will be swinging back to purpose built CPUs that have embedded vector and matrix capabilities, highly tuned for specific algorithms. This will be specifically true for intermediate edge computing and endpoint IoT devices that need to compute locally because shipping data back to be processed in a datacenter doesn’t make sense at all, either technically or economically. [...]
“With over eight billion voice assistive devices. We need speech recognition on sub-$1 microcontrollers. Processing everything on server just doesn’t work, physically or financially. Cloud computing bandwidth aren’t free and recognition on device is the only way. A voice activated coffee maker using cloud services used ten times a day would cost the device maker around $15 per year per appliance. Computing ML on device also benefits latency, reliability and, crucially, security.”
There’s also some very interesting security improvements, which ARM broadly calls “Confidential Compute Architecture (CCA)”. The idea is that some data or code can be shielded from any modification, even from OS-level access, via doing the compute in an hardware-protected area on chip. Cool stuff.
‘Intel plans to change its [nanometer] numbering conventions to match the industry standard’
Within the semiconductor industry, though, the divergence in numbering has put Intel at a distinct disadvantage.
At nearly every technical presentation, Intel engineers take pains to argue that its chips perform every bit as well as those its competitors describe as belonging to a smaller generation.
Intel now appears to be tiring of the argument.
Employees say Ann Kelleher, the Hillsboro vice president who runs Intel’s manufacturing group, notified them last week that Intel plans to change its numbering conventions to match the industry standard. (Source)
These marketing numbers shouldn’t matter, but they do. The shoe used to be on the other foot when Intel was on the netburst architecture and could crank up clock speeds really high on its Pentium 4 chips, and AMD had much shorter pipelines and lower clock speeds.
AMD tried explaining for a while that performance was similar at lower clock speeds, but after a while gave up and just added “Intel equivalent” numbers to its chips, so you had an Athlon clocked at 1.8ghz that may be branded as a “2800+” or whatever to show how it could compete with a Pentium IV at 2.8ghz (I’m just making up these numbers to illustrate).
h/t Dylan522p (522p? Couldn’t go all the way to at least 1080p, Dylan? Maybe 4k at some point?)
TSMC’s 4nm node to “enter risk production in Q4”
TSMC will move N4 (namely 4nm process) to volume production in the fourth quarter of 2021, ahead of the 2022 timeframe set previously, according to sources at fab toolmakers. [...]
The foundry house's N4 - an extension of its 5nm process family - will be available for volume production in fourth-quarter 2021 ahead of its original scheduled (Source)
Apparently, Apple has already booked the initial N4 capacity for its ARM Macs (I’m guessing capacity will be too low at first to go in the next iPhone, but may be high enough for Mac..?).
Jupiter and its Moons dancing with our Moon
In this picture, taken in July of 2013, we [an] occultation—the waning moon, Jupiter, and the four Galilean moons: Callisto, Ganymede, Io, and Europa.
The kind of photo that you probably can’t take very often. It must be really frustrating if you plan it all months in advance, set up all your gear, and then it’s really cloudy that night...
Photo by Juerg Alean. Source. Via Massimo
Turns out tiger stripes are also on their skin…
Best comment on Reddit:
I didn't read the title and I thought I was looking at an aerial view of a field of dry grass after a brushfire.
The Arts & History
Woke Up to Ducks in My Pool — Felt Like Tony!
Beethoven’s Deafness Increased his Originality?
By the age 45, he was completely deaf. He considered suicide, one friend reported, but was held back only by the force of “moral rectitude.”
It’s here that Beethoven’s story veers toward legend. Cut off from the world of sound around him, working only with musical structures dancing through his imagination, at times holding a pencil in his mouth against his piano’s soundboard to feel the consonance of his chords, Beethoven produced the best music of his career, culminating in his incomparable Ninth Symphony, a composition so daringly new that it reinvented classical musical altogether.
“It seems a mystery that Beethoven became more original and brilliant as a composer in inverse proportion to his ability to hear,” writes Brooks. “But maybe it isn’t so surprising.”
As Brooks elaborates, Beethoven’s diminished hearing limited the influence of “prevailing compositional fashions.” Whereas his earlier work was “pleasantly reminiscent” of his instructor, Josef Haydn, his later work was spectacularly innovative. “Deafness freed Beethoven as a composer because he no longer had society’s soundtrack in his ears.” (Source)
It’s an interesting theory. I don’t know if it’s true, and there’s probably no way to know, but I think there’s an interesting meta-point about tuning out the noise and not being too influenced by others.
Vincent Van Gogh Time Travels, Sees His Work in Gallery
Yesterday was Vincent Van Gogh’s birthday, and for the occasion, here’s a nice scene from Doctor Who:
Vincent Van Gogh Visits the Gallery (3 minutes)
I’ve never actually watched Doctor Who, but Sarah McGonagall shared the scene on Twitter, and I thought it was really nice and wholesome.
I think you've discovered what flow is like when you're creating something :).