Discover more from Liberty’s Highlights
149: Jobi Lütkos, Benevolent Neglect, Twitter's Platform Move, DuckDuckGo's Rise, Satya Nadella on Microsoft, SpaceX Starlink, Nvidia + Google, and Mescaline Medication
“Windows, we could just rename it Azure Edge"
If it costs you your peace, it's too expensive.
🛀 I was thinking about how what I’ll call here the ‘reflexive contrarian instinct’ or the ‘hipster reflex’ is a lot less valuable today than it was when I was growing up, in the pre-to-early-internet world.
The reason is simple: If what is mainstream/easy-to-find had to go through very few gatekeepers, then there’s a pretty good chance that those gatekeepers have missed a lot of what is good, or at least, that the structure of things means that anything that is niche won’t surface on the very few distribution channels that have to serve the big bulge at the center of the bell curve.
But today, what makes it to the surface hasn’t just been picked by a few editors or printing press heirs or whatever. It often emerges after being uncovered, vetted, and amplified by thousands or millions of people, many of whom are experts at their own little niches.
A lot more of the good stuff gets found, and there’s many more big distribution channels that — while they aren’t as high market-share as the few of pre-internet — are big enough to put things on the map.
In short, being a mainstream act back then meant that you had middle-of-the-road mass appeal and got through a few gatekeepers.
Being mainstream today can also mean that, but it could also mean you got traction in some niche and either got big enough there to have crossover success or something of the sort, enough to be known by tens of millions of people.
To go back to that hipster instinct, back then it helped you discover a bunch of good underground stuff that would never make it to the mainstream.
Today, it can still do that, but it also pushes you away from a lot of good stuff that a few years ago would’ve stayed underground but that today found a distribution channel thanks to the decentralized emergent discovery mechanism of the internet.
Some song goes viral in a TikTok video that no gatekeeper approved, and suddenly millions of people know it — what was the equivalent 20-30 years ago?.
(I guess all my examples are about music because that’s what I was thinking of when I had this idea, but it applies to many other formats too)
🗣 How many close friends do you have?
15% of men and 10% of women in America have no close friends. The decline is most pronounced among men. In 1990, only 3% of men reported having no close friends. In 2021, the number rose to 15%. We need to help; friendship strongly predicts life satisfaction.
🐿 🚗 Driving back from daycare, I saw a car run over a chipmunk. The poor thing was on the median, freaking out, panicked and ran straight into the back wheel...
I wonder if mature mesh-networked self-driving cars will someday have such good sensors & millisecond-control that they'll save lots of wildlife too. I’m not talking today or tomorrow, but when the tech is mature and all vehicles on the road are basically robots, removing the human element. And I’m also not saying “at any cost” (swerving hard or whatever), but in most cases, you can probably do it just by slowing down or changing direction a bit.
If they’re all communicating at very low latencies, it's trivial to make micro-adjustments to avoid animals, and all other cars around would know what your car is planning on doing before it does it.
To a computer, it's all happening in super-super-slow motion.
I’m not saying it should be a top priority, but it’s easy to imagine that once you’re really really good at all the other stuff — that someone walking in traffic on purpose would have a hard time being hit — that you would then start to work on other problems too. At first you’d focus on deers and larger animals that can injure or kill passengers, but eventually, why not try to save birds and squirrels and groundhogs or whatever?
💚 🥃 I figure that the price of a couple coffees or one alcoholic drink isn't a bad trade for 12 emails per month full of eclectic highlights and investing/tech analysis.
The entertainment has to be worth something, but for those that care most about the bottom line:
If you make just one good investment decision per year because of something you learn here (or avoid one bad decision — don’t forget the value of *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 — if you don’t see the support plans, just make sure you’re logged in substack, top right):
Investing & Business
Benevolent Neglect (what about it? 🤔)
I think the concept of “benevolent neglect” is often overlooked and underestimated.
What I mean is when someone or some entity has de facto control of something, and they could make all kinds of moves that would advantage them (at least in the short-term), but they don’t, they just leave it alone, allowing bottoms up, emergent creativity to try lots of things and create an interesting open ecosystem.
The clearest example I can think of is Apple with podcasting.
They basically created the modern version of the medium; it’s named after the iPod and the iTunes directory is still the canonical “podcast DNS table”.
But they mostly didn’t do anything with the influence they had on the space for a loooong time. They could’ve tried to lock it down, to create proprietary formats with embedded analytics, ways to track and monetize things or prevent competitors from accessing their directory, etc.
Instead, a great open ecosystem sprung up from a thousand seeds planted by anyone who felt like gardening on that plot of land. Many didn’t get anywhere, but some became big, famous, rich names, yet have never paid a cent to Apple for the platform.
Don’t get me wrong, Apple made a ton of money from podcasting, but it was indirect, by selling more iPods and iPhones. Podcasts made the devices more sticky and valuable to users, and the experience was better on iOS than Android for a long time (probably still is — whenever I saw stats with app/platform breakdowns, iOS punched way above its general marketshare weight in podcasting).
What are some other examples of benevolent neglect?
A lot of the early internet probably falls in that category.
Those in charge just didn’t see it as their purview, or were ideologically opposed to trying to lock down, control, and monetize the web, so this allowed a crazy amount of innovation built on top of open protocols.
Imagine if the email protocol was owned by a single company, the way Twitter owns what is basically a protocol (and how little innovation that protocol saw for most of its existence), or if a single company had owned the web (ie. can’t just host your own website, have to build it inside a walled garden like Facebook, or AOL back then).
I think that would’ve been a tragedy.
🐦 Twitter’s Platform Future? (3P Distribution/Partnerships)
Twitter is at a crossroads. Well, everyday’s a crossroads, but y’know what I mean.
They can decide that Substack is the competition and that they need to go after them with Revue, and they need to find ways to use their distribution and integration muscle to favor Revue and make Substack a second-class citizen, hoping that enough creators (who are generally Twitter power-users) will switch…
Or they can put on their “platform” hat and sit down with Substack over free range, grass fed steaks and make a deal.
I’m envisioning something like:
Signups are integrated, one-click from Twitter bio button, can do paid subs via credit card that is on file at Twitter, etc. Twitter would take a cut of revenue as a funnel/distribution channel (which is what they already are, but unpaid).
This way, they don’t have to try to do everything themselves (often not as well as more focused teams) and end up having to fight everyone and create a worse experience for users. I bet this platform approach would make them more money than just Revue on its own (keep Revue, but just have it as one more option on the menu).
Similar to how Amazon figured that Prime Video made money by distributing other paid channels and taking a cut. That’s high margin revenue. Others are building and maintaining the product, you just have to worry about being the best platform/distribution.
Of course, the turd in the punchbowl is the even bigger platforms (G and A) taking big cuts on these types of digital transactions, but I’m hoping that this will change soon and will allow these types of business models to operate without a huge cut going upstairs.
Now replace “Substack” in the above with “OnlyFans” or “Fanhouse” or “Patreon” or whatever.
DuckDuckGo Search Engine’s Gone Parabolic
The duck company published an update on how its doing, and it’s pretty impressive for a company with 129 employees:
DuckDuckGrowth over the last 12 months: 50M+ app downloads, 55% search traffic increase, #2 search engine on mobile in the U.S., Canada, Australia, the Netherlands & more.
In the coming weeks, we're adding new free privacy protections in beta including a cross-platform email privacy solution & app tracker blocking on Android devices.
Later this year, we'll release a beta of the first-ever DuckDuckGo desktop app, which can be used as a primary browser.
DuckDuckGo has been profitable since 2014 with revenues now $100M+/yr. In late 2020, we completed a $100M+ mainly secondary investment from new & existing investors.
How many users do they have? Ha! That’s hard to tell for a privacy-centric company:
We don’t track our users so we can’t say for sure how many we have, but based on market share estimates, download numbers, and national surveys, we believe there are between 70-100 million DuckDuckGo users.
Their mobile app, which is a combined search app and browser, was the most downloaded browser on Android in Q1 and 2nd behind Chrome on iOS. They’re projecting about 75 million app & extension installs in 2021.
It reminds me a bit of the early days of Firefox (then known as Firebird), and how it came out of nowhere and eventually did real damage to Internet Explorer’s dominant market share because it provided a better experience and spread quickly through word-of-mouth1.
I’m not saying that DDG will do this kind of damage to Google’s market share — it’s a lot more entrenched than IE was — but I think the duck can built itself a really strong niche business if it keeps doing what it’s been doing.
Who knows, Apple may even try to acquire them soon…
(saw photos of Tobi and Jeff side by side, and Tobi wasn’t wearing his traditional beret, and I couldn’t help myself…)
h/t Friend-of-the-show Fahd Ananta
App Store & Newsletter Fee Wars
Speaking of Shopify, Harley Finkelstein just fired a new shot in the great app store fee war of 2021:
To our partners: on Aug 1st we are removing all revenue share on your first $1M. That means you will keep 100% of your first $1M when you sell on @Shopify’s app store.
The best part? At the end of every year, the numbers reset. Every single year, your 1st million is all yours.
And then Zuckerberg announced during the unveiling of their newsletter product:
Unlike other newsletters, Bulletin will not take a cut of the earnings [at launch]
I guess they had to pull a dramatic move because all the written-word creators are on Twitter and not on FB.
Satya Nadella Interview (Microsoft CEO)
I like this line from a few years ago:
“Windows, we could just rename it Azure Edge, if people would let us. It’s just an extension of our big business, which is Azure. We’re in a mobile-first, cloud-first world.”
Nadella updates the thought:
There are multiple cloud providers and a lot of cloud computing. So anything that is a client operating system ultimately does rendezvous with cloud computing. In that sense, technically and business-model-wise and usage and experience-wise, it’s the cloud and the edge.
The billion-plus users of Windows, for sure, we think of from a distributed computing architecture perspective as the edge of Azure. And you could even say the Windows folks would look at Azure as the cloud for Windows. So that’s, I think, absolutely right.
I like his view of how Windows needs to play nice with the mobile platforms, which hasn’t always been Microsoft’s view:
the reality is any Windows user — we have to start with the assumption that they have a phone and that phone may be Android and iOS and we have to design for it.
Some platform 🔥 clearly targeted at (mostly) Apple and Google:
In our case at Microsoft, I’ve always felt that, at least the definition of a platform is: if something bigger than the platform can’t be born, then it’s not a platform.
The web, it grew up on Windows. Think about it. If we said, “All of commerce is only mediated through us,” Amazon couldn’t exist, if we had somehow said, “We’re going to have our own commerce model.”
You can read the whole thing here.
Science & Technology
Ever seen coffee fruit?
Recently realized that I didn’t have a good mental image of what the coffee fruit looked like, before it was turned into the roasted beans that we’re grinding in our Baratza, and neither did my wife.
So did the only responsible thing to do in such a case: Looked it up. I figure many of you probably were missing a good mental image of the coffee fruit, so here you go.
“Live” View of SpaceX Starlink Satellites
Mescaline Enters the #SchroomBoom Chat
The use of mescaline, a psychedelic substance derived from the peyote cactus, appears to be associated with improvements in depression and anxiety, according to a new preliminary study published in the journal ACS Pharmacology & Translational Science. (Source)
It’s not the best study from a methodology standpoint, as it’s retrospective and self-reported, but you have to start somewhere and get those promising signals to be able to fund and justify more rigorous studies. Looking forward to more data on this in the coming years…
Nvidia’s Getting Closer to the Edge
After a recent deal with Cloudflare to put GPUs on its edge platform, the green company (not that one, the other) is partnering with Google Cloud:
NVIDIA today announced that it is partnering with Google Cloud to establish the industry’s first AI-on-5G Innovation Lab, enabling network infrastructure players and AI software partners to develop, test and adopt solutions that will help accelerate the creation of smart cities, smart factories and other advanced 5G and AI applications. [...]
NVIDIA announced in April it is teaming with Google Cloud, Fujitsu, Mavenir, Radisys and Wind River to develop solutions for NVIDIA’s AI-on-5G platform. Google Cloud is extending the Anthos application platform to the network edge, allowing telecommunications service providers and enterprises to rapidly deliver new services and applications at the 5G edge. (Source)
Signal & Noise vs Aha! Moments
Great stuff by Venkatesh Rao aka VGR. Quoting a lot because there’s a lot that’s good:
Have you noticed that despite the supposed information overload problem the incidence rate of your “how did I not know this!” flow of input-Aha moments has been slowly increasing over the last few years. Not signal-to-noise ratio, actual absolute signal levels? Mine has.
Signal to noise (SNR) is actually a bad measure for intelligent systems. You’re not a simple amplifier or something. Noise should be weighted by how costly it is to tune out, and signal by how costly it is to tune in. Most noise is very cheap to cut out.
If you have clarity around what you care about, 90% of noise is filtered out by your unconscious. It doesn’t even register. You have to learn meditation to even notice it.
Of the remaining 10%, 9% is classification errors caused by your own clarity issues. 1% is real noise. [...]
All human signal/noise filter problems are generalizations of picking out your name being mentioned in a crowded room with lots of conversations. Salience = identity = interest = attention.
Know thyself = know your True Name better = be Very Online better
I think it was @dhh who said something like burnout is due to being underpurposed rather than overworked. Tracks for me. Similar thing here. It’s not a filter failure problem. It is desire imprecision problem.
Counterintuitive but… You don’t get to accurate and precise desiring via being accurate but imprecise and narrowing the focus gradually. You get there by being precise but inaccurate and retargeting iteratively. [...]
The Arts & History
Music to Read/Work to: Tingvall Trio
A long time ago, I noticed that if I tried working while listening to music with vocals or catchy choruses, it was distracting me from what I was reading or thinking.
It’s not a deep insight, but I figure that if I had to discover it, there’s probably people out there who haven’t yet noticed consciously.
After figuring that, I got pretty deep into some of the more instrumental genres, like jazz, classical, electronic, ambient, etc.
The album I want to share today is from a trio with a Swedish pianist, a Cuban bassist and a German drummer. I tend to love these piano-centric trios and quartets for work, maybe because instruments like the saxophone and trumpets are closer to the human voice and it triggers the language part of the brain..?
Anyway, to me this is beautiful music, fairly airy and relaxing, but with just enough depth and layers to be interesting over multiple listens and at varying levels of focus.
Here are convenient links:
Villeneuve’s ‘Dune’ Pushed Back to October 22nd, 2021
My man Denis.
Looking forward to this one, I hope it’s as good as his other films, and that he gets to make the sequel, unlike with Bladerunner 2049 (which I saw twice in theater — now *that* was a big screen movie).
Conversely, it’s possible to imagine a very different trajectory for the web if Microsoft had somehow been able to keep being dominant with IE and had successfully locked competitors out by using even more closed-standards than they did (ActiveX and such).