124: Unexpectedly Moderna'ed, Hyperscale Cloud Ecstasy, Unfreezing the Data Lake, Microsoft Store, Spotify, Starcraft, Autonomous Vehicle Ecosystem, and Impostor Syndrome
124: Unexpectedly Moderna'ed, Hyperscale Cloud Ecstasy, Unfreezing the Data Lake, Microsoft Store, Spotify, Starcraft, Autonomous Vehicle Ecosystem, and Impostor Syndrome
"it's a highlight reel of a highlight reel of a highlight reel of a highlight reel"
If you can, build it. If you can’t build it, fund it. If you can’t fund it, champion it. If you won’t champion it, at least stay out of the way.
—John Carmack
🤘 So yesterday was a good day!
I got an early birthday present when a high-school friend told me that a clinic had 40 doses that they didn’t have appointments for, and would have to throw away (😱) at the end of the day. So if I was interested, I could come in, along with anyone else I could find.
My wife and I got the shot, plus a bunch of family and friends that I was able to reach on short-notice.
I heard from a friend who was there after I left that they had given 38 of the orphan doses, and asked people in the room if they knew anyone else who might need one, and two people raised their hands and called people. So hopefully all the doses were used.
Best day in a long time! Finally a concrete marker of the beginning of the end for this pandemic that is truly starting to grind me down…
📚 Follow-up to the ongoing meta-book/reading discussion:
I’ve written about book-pairs in edition #115, and about how I read multiple-books in parallel to increase my chances of having something — anything — that I’m motivated to read at any time in edition #122.
Friend-of-the-show and 𝕖𝕩𝕥𝕣𝕒-𝕕𝕖𝕝𝕦𝕩𝕖 supporter (💚💚💚💚💚🥃) Joseph Hodes emailed me something quite good by Slava Akhmechet that takes the “book pair” concept to “book cluster” territory and adds some more dimensions to the selection:
A single book is a pinhole view of the world set up by the author. You have no input into its contents, and therefore cannot change the orientation of this view. But you do choose the books you select. That means you can stitch together multiple pinhole views into a unique lens to examine the world— one that no one else will have unless they use the same list of books to stitch together the same lens.
That’s some good writing. Back to Slava:
For example, you can look at the world through history of technology that became ubiquitous. Here is one possible list of books to stitch together this lens: The Victorian Internet, Empires of Light, The Wright Brothers, The Network, Hackers. Very few people in the world read all five of these books. Even within Silicon Valley, where everyone's living depends on creating new ubiquitous technology, your understanding of how technology becomes ubiquitous will be in the 95th percentile (and likely higher) if you read these five. You will now be armed with a unique instrument that few others possess, and assuming twenty books per year, it only took you three months to acquire. [...]
If you organize your reading this way, your bookshelf won't be arranged by genre like a typical bookstore. Rather than having sections for e.g. biographies, essays, novels, military history, all sorted by author, you will instead have sections for addressing different problems, likely sorted by clusters of books in order that you encountered them. The arrangement won't make sense to anyone except you.
I think it’s kind of something I’ve been doing at times, but not top-down, in a designed way, but just because I had high-enough interest on a topic to sustain the interest to see it from multiple angles (f.ex. war memoirs and high-level overviews, multiple conflicts, etc).
But I like the concept enough that maybe I should try to harness it in a more deliberate way.
Anyone else has been reading books by clusters centered around a problem/specific thing they want to learn about from various angles to paint a richer tapestry inside the ol’ upstairs gallery?
🤔 What is the best way to teach/train someone to be more comfortable with ambiguity?
📝 Dammit, I'm becoming increasingly convinced that I want to try the Roam Research's note app…
But I don't need more note-taking software in my life 😩 I already write stuff in four apps (Notion, Apple Notes, Apple Pages, and Google Docs), and I don't think Roam really replaces any of them, would be more additive.
Why am I interested in Roam?
Most of my current note apps are more of the "architect" model, where you need to create the organizational structure and then slot content into the right places.
Roam would fill my need for that "gardener" part where structure emerges over time as connections build up and various nodes in the graph grow with use.
It’s a bit more like a brain, where neuron myelination happens as pathways get use, creating stronger and more complex memories and knowledge clusters.
If you want to learn more to see if Roam is for you, this is probably the best intro video that I’ve seen (and this is pretty good too, to show the kind of stuff you can do with it).
I’m sure I’ll write more about it later, including some of the alternatives that are designed around the zettelkasten method, like Obsidian, Athens Research, Logseq, etc. Stay tuned.
Update: Not long after writing the above, I saw this video and it half-convinced me that Obsidian may be a better fit for me than Roam.
It at least convinced me to give both a try and compare them for my own use case.
The most intriguing aspect is that since Obsidian is less outliner-based than Roam, it may actually be what I need to fully replace Apple Pages (if you’re not familiar with it, it’s basically like MS Word in the way I use it) as the software I use to keep my investing journal, and I could use a bunch of Obsidian's features to make this journal even more useful (more powerful search, more powerful ways to link things bi-directionally and graph those connections, replacing my reverse-chronological linear model with a multi-document approach by topic/company while still keep a central trunk with chronology, etc).
So I haven't decided anything yet, but this mini-project just got more complex.
Oh well, nobody said building and managing a knowledge base was supposed to be easy. I certainly can't do it as effectively just in my brain as I can with a note system, and I can't do it as well with a simple plain-text note system as I can with these next generation note apps, so it's probably worth it.
The right tool for the right job.
🛀 🧬 I wonder how instinct is encoded in DNA.
If we agree that animals have it, some in very obvious ways from birth, then there must be atoms that encode it (ACGT nucleotide bases pairs, codons, etc) on the DNA strand.
This means that, in theory, it should be possible to find, reverse-engineer, and modify it.
We could have a per-species translation software that allowed us to read and write instinct.
We could program a gazelle to start moon walking moments after birth, or rodents to be afraid of human houses, or things of the sort.
This premise would probably make for an interesting sci-fi short story…
☃️ I just realized that someone who sells some Intercontinental Exchange to buy some Snowflake is basically trading ICE for SNOW1.
💚 🥃 Someone who really really knows about newsletters recently became a paid supporter (💚) and said something like: “People are happy to pay for something, but they don’t want to feel like they’re doing charity.”
I gotta admit, that cut deep.
I do think it’s a little annoying that with my current model, I gotta constantly do that Wikipedia-style banner “hey, help support this project!”.
Paywalls do get over the bystander effect where everybody goes: “well, others will pay and it’ll be fine, so I don’t need to do anything” (and then everybody thinks the same thing, so nobody pays and the thing goes away).
Paywalls are very clear: You pay this, you get that.
I certainly don’t want you to feel like I’m asking for charity. I’m trying to put as much value as I can out there, and just capture back a bit of it… This is a two-way exchange, not one-way. Only pay if you feel like you’re getting value.
But paywalls also mean that most people can’t read, so editions that took just as long to write would only reach 5-10% of the mailing list, and can’t be shared online. It probably means less feedback from you guys & gal and fewer interesting conversations, which kinda sucks.
So there are downsides too.
That’s why I’m trying this weird “free + support” model to try to find a balance somewhere in the middle, where I can do this sustainably and everyone can join the party, whether students or in a low-wage country or whatever.
Right now we’re at 3.3% paying supporters and 96.7% free readers. I think things could work at 5%/95%, so if you want to move that needle, it only takes 20 seconds and means a lot to me:
Investing & Business
The Ecstasy of Owning a Hyperscale Cloud
I recently saw this quote from a Bernstein report about Microsoft:
As MSFT improves margins for its cloud platform (Azure) the margins for *ALL* its cloud offerings improve
It’s fairly obvious when you think about it, but I hadn’t quite thought about it explicitly before, and when I did, I realized the many ways in which this was advantageous to the giant cloud platforms.
First, selling access to an hyperscale cloud is a good business in itself.
This is what a lot of the media focuses on, renting servers, the IaaS layer. But there’s plenty of layers where the value is more from the software side (PaaS and SaaS), and that can be a much better business, especially when you take into account the vertical integration with the underlying infrastructure.
For example, let’s say that Microsoft datacenters get more efficient. They find ways to use less power, get better prices from suppliers because of increasing scale, tweak some routing, ML, and compression algorithms and make them more efficient, etc.
These benefits can result in better margins (lower costs = high margins, all else equal — duh). But how they “spend” that margin, and where else the benefits show up is the interesting part to me.
On one hand, they can decide to maintain their pricing and pocket the higher margins. Or they can cut prices and “buy” some growth, or at least maintain competitive position vs others who may also be cutting their prices (but by the time you reach hyperscale, only the biggest competitors can follow you down the cost curve, at least when it comes to things where scale is the primary benefit).
But there’s also the benefit to whatever runs on top, and that’s the cool part.
So for Microsoft, if they improve margins on Azure, it also helps all their own stuff that runs on Azure, like Office 365 or even LinkedIn. All these already very-profitable businesses get the automatic profitability/performance lift because they’re built on top of the same platform.
It also gives them a big advantage when competing with third parties that don’t own a ginormous cloud.
When AWS is competing with Elastic, Elastic has to pay for its infrastructure at some “retail” price (which includes a profit margin for AWS/Azure/GCP), while AWS is running its Elastic clone/fork on infrastructure where they don’t really have to pay for the profit margin part (or at least, if they do for accounting reasons, once everything gets consolidated in the end, they don’t really).
So when Microsoft buys Nuance and keeps it running on Azure as they did before the deal, the economics of Nuance automatically change for the better because the New Nuance inside of Microsoft is in fact paying a much lower bill for its cloud infrastructure than it did as a separate company.
It’s a dynamic similar to Apple making its own CPUs (packaged in a SoC as the M1). They don’t have to pay for Intel’s margin for the CPU, so they can either price their computers lower (all-else equal) and sell more of them, or pocket that margin (since they are at a scale where the R&D costs to make a CPU can easily be absorbed). Likely something in between.
Good position to be in, no? AMZN 0.00%↑ GOOG 0.00%↑ MSFT 0.00%↑
Unfreezing the Data Lake 🥶
Maybe I’ve got Snowflake ❄ on the brain (is that like a brain freeze?), but was thinking that as the ‘data cloud’ matures & these two-sided data marketplaces grow, companies with quality* data asset may see those assets become incrementally more valuable because it’ll be easier to find buyers/for those buyers to use all that data.
For example, S&P Global (with the newly purchased IHS Markit data) may have a lot of things to sell there, and it may be lower friction than selling through clunkier, higher-CAC traditional channels.
* The word "quality" in front of "data assets" matters a lot here. If you have something unique, it'll likely be more valuable. If you have the same thing as everyone else, it'll become less valuable because buyers will be able to more easily price-compare.
‘Verizon sells Yahoo and AOL businesses to Apollo for $5 billion’
What if Yahoo had sold to Microsoft when Ballmer offered $44.6 billion in 2008 (and what would that Microsoft stock be worth today, as I'm assuming it wasn't just a cash offer — and Ballmer probably would’ve sold the Alibaba stake within a year or two, like Microsoft didn’t hang to its Apple and Facebook stock as far as I know). I realize what Apollo bought isn’t quite the same assets, but still…
‘Microsoft reducing Windows store cut to 12 percent’
he software giant is reducing its cut from 30 percent to just 12 percent from August 1st, in a clear bid to compete with Steam and entice developers and studios to bring more PC games to its Microsoft Store. [...]
These changes will only affect PC games and not Xbox console games in Microsoft’s store. While Microsoft hasn’t explained why it’s not reducing the 30 percent it takes on Xbox game sales, it’s likely because the console business model is entirely different to PC. Microsoft, Sony, and Nintendo subsidize hardware to make consoles more affordable, and offer marketing deals in return for a 30 percent cut on software sales. [...]
Valve still takes a 30 percent cut on sales in its Steam store, which is reduced to 25 percent when sales hit $10 million, and then 20 percent for every sale after $50 million.
Valve has to be minting so much pure margin… I wish they were a public co just so I could see those 95% FCF margins (or whatever). MSFT 0.00%↑
Interview: Aaron Edelheit
Bill Brewster had a very interesting conversation with friend-of-the-show and supporter (💚 🥃) Aaron Edelheit.
Mindset Mastery (podcast)
I have no pre-existing interest in cannabis on the investment side, though I've been following the science a bit (Peter Attia’s podcast on THC/CBD was very interesting), but I learned many things from Aaron and it helped me realize that the space isn't just the crazy Robinhood stocks that I saw in my peripheral vision these past few years.
There's some interesting dynamics at play, and possible barriers to entry for some of the US operations.
Doubt I'll suddenly invest in the space — I don't know enough about it and am not particularly attracted to studying it — but learning new things is its own reward, and who knows where it’ll lead over time.
Here’s something Aaron wrote on the topic:
Back to the pod: The part near the end where Aaron talks about his workaholism, and quantity of work vs quality, about burnout, mental health, work life balance (as cliché and meaningless as the phrase has now become... We need a new one) also resonated with me.
I'm not a workaholic, but I've had to optimize things a lot in recent years, in large part because having kids changes everything in someone's life, and in the past year because of the pandemic.
Exponent on Podcasting & Spotify (Podcast)
Good episode, mostly about podcasting in general, and Spotify’s recent moves:
Ben Thompson and James Allworth make many good points.
Generally, I think there's a lot of good stuff happening in podcasting. I love seeing the vibrance of the ecosystem, especially on the creator side. I think there’s also some bad stuff starting to bubble up, as I’ve discussed previously.
I wish we'd find ways to keep the good and minimize the bad a little more, but the die hasn't been cast, so there's still a chance, especially with Spotify announcing that they'll implement their paid subs in a way that is friendly to third-party players and allows creators to own their customer relationships.
What is a third-party player worth to me?
Let’s see: Overcast has a dynamic silence-shortening feature that Apple and Spotify's players don't have.
So far, since I’ve been using the app (summer of 2014), that feature alone has saved me about 639 hours of my life (there’s a clever counter that keeps track on the setting page in the app, though I’m not even sure if it has been keeping track since day one, so my number may actually be higher).
So either I've listened to the same amount of content and had 639 hours to do other things, or I've heard 639 hours of extra information and/or entertainment (I know myself, it’s the latter).
How much do you value an hour of your life? Whatever price you put on it, multiplied by 639, it's a significant number for me...
And that’s just one aspect of the app that I prefer, there are others, though they’re harder to quantify.
So you can see why I’m quite invested in the podcasting ecosystem not turning into walled gardens that force you to use a couple of big players designed for the “average listener”, and take it or leave it. *insert William Wallace ‘freedom’ scream* SPOT 0.00%↑
What Starcraft Teaches
Good line in a email by friend-of-the-show Jeremy Saltzberg (reproduced with permission):
Starcraft is a wonderful game for teaching mental models such as resource optimization, competitive strategy, and opportunity cost.
You can see why Tobi of Shopify likes it so much.
Or why Factorio is also a Tobi favorite (I’ve never played that one, but if it had been out when I was growing up, I certainly would’ve been obsessed with it for a while).
Science & Technology
Mapping the Autonomous Vehicle Ecosystem
Via Granite Capital (they have a Substack too):
With Lyft selling their AV unit to Toyota we are starting to see a gradual shift of R&D intensive AV IP to a handful of leaders like Nvidia, Amazon, Hyundai, GM, VW, Google with the balance sheet muscle to continue developing the tech
Nvidia seems to have R&D’ed itself to a pretty central spot in this jungle.
h/t friend-of-the-show Muji
Space Debris Simulation from Two Satellites Colliding
This gives an idea of the kind of mess that a large-scale war where sat killer weapons are used could make up there.
We already have a lot of stuff floating around and lots of debris (basically bullets going at hyper-speeds), and cleaning it up is one of the biggest challenges for the space agencies.
The Arts & History
Porter Robinson on Impostor Syndrome
I dig Robinson’s 2014 album ‘Worlds’. He just released a new album, but I’ve only heard about half of it once, so I don’t have an opinion on it yet.
Porter did an AMA on Reddit to promote this launch, and I thought what he said in reply to a question about impostor syndrome was interesting:
you're comparing your insides to everyone else's outsides
comparison is a dangerous game anyway, but it's also almost impossible to do with any degree of accuracy or real perspective.
i talk about instagram this way a lot: in your feed, you see other people's photos after they've posed with the intention of being photographed, with selected lighting, they've taken 400 photos and chosen the most flattering one, and then adjusted colors and often times used beauty apps. and there's another degree of "filtration" there -- sometimes people take all of those steps and then don't post a photo.
it's a highlight reel of a highlight reel of a highlight reel of a highlight reel.
and then you compare that to your own face when it accidentally pops up in the forward-facing camera. lol
making music is like that. artists tend to show the best 5% of their work. that's why it's better to have compassion for yourself as you're being creative rather than judging the thing you're doing. that's your insides YA GOTTA LOVE IT
🤘
I’ll show myself out…