347: Ben Thompson's Pivot, Elon Musk + Twitter, Hyperscalers, Spotify Focus, Facebook Sentiment, and UV-C Lights
"It may get easier to have a really unique ugly sweater"
We judge ourselves by what we feel capable of doing, while others judge us by what we have already done. —Henry Wadsworth Longfellow
🗣🎙♾ 🔁 Yesterday, I recorded my second podcast ep on Infinite Loops (♾ 🔁) with friend-of-the-show and supporter (💚 🥃) Jim O’Shaughnessy. It’ll be released on November 24th.
In the meantime, if you haven’t heard it and want to check out my first appearance on the show, you can listen here:
🤔💭🗺 🎮 Doom may be a 30-year-old game, but the open-source game engine has been continuously updated and there’s still a vibrant map-making community. (I post some of my recorded games here)
One really interesting genre is ‘maps made from memory’. Some people have re-created the original Doom and Doom II maps entirely from memory, without looking at the originals.
When you play through them, everything has a weird, surreal quality. Almost everything is how you yourself would remember it (if you’ve played the originals hundreds of times like us Doom geeks…), but distorted and not quite right. Corridors are longers, doors wider, rooms don’t quite feel right, some walls aren’t the right colors, etc.
It’s like being in your childhood house *in a dream*.
This made me think that it would be interesting to do something similar with other mediums.
What about recreating things like short stories, novels, films paintings, sculptures, songs, etc, entirely from memory without using the original as a reference? Or levels in other games (Portal? old Zelda games?) 🤔
That could be interesting…
👂 Animated GIFs that don’t have audio teach us something about the human auditory cortex, because when they’re for a film scene that I know well, I "phantom hear" the audio from that scene in my mind.
A common line in neuroscience is “neurons that fire together wire together”, and clearly visual memory is linked with auditory memory.
💚 🥃 🍄 You’re Super Mario (ok, you can be Luigi if you prefer) and this newsletter is your source of power-up mushrooms. Keep ‘em coming.
Liberty’s Highlights is reader-supported. To support my work, consider becoming a paid supporter. 🍄
📊 A Word From Our Sponsor: Koyfin 🕵️♂️📈
Koyfin gives you professional-grade data and analytics with a user-interface from the 21st century to research companies like never before. New features are frequently added, so it gets better all the time!
Quickly stay on top of what's happening in your portfolio and understand how the market is changing with Koyfin. 🕵️♂️📈
And because you’re awesome, you deserve a discount:
*Offer only valid for new users
🏦 💰 Liberty Capital 💳 💴
Hyperscalers growth rate over time ☁️☁️☁️☁️
📝 Ben Thompson’s Slow Pivot 🎙
With success comes possibilities.
A lot of those you should say no to, because they take focus away from doing the thing that you’re best at, that has gotten you to be successful in the first place.
The most common mistakes after you make it are to either spread yourself too thin, which makes the quality of your ‘main thing’ suffer, helping competitors catch up, or to stay focused, but on the wrong thing, like maybe some new ‘shiny object’ that seemed more desirable than the original ‘main thing’ but just never quite took off (it’s hard to capture lightning in a bottle more than once).
I think Ben Thompson (💚 🥃 🎩) is a good example of someone building on success the right way. It’s not surprising because his shtick is business strategy and clearly he’s good at analyzing his own business. It’s worth examining what he’s doing to see what we can learn.
What am I talking about?
Here’s what I’m seeing: When Ben started out, as he mentioned in recent podcasts, nobody knew who he was and he couldn’t get an interview at Google (where they now recommend his works). The best use of his time/energy was clearly to write and publish analysis because it didn’t require the approval of any gatekeepers or industry access.
After years of sticking with it and doing really good work, he got to a point where he now has as much access to the people he’s writing about as anyone else in the industry, and the dynamic with gatekeepers has inversed — Ben is now the one with distribution power!
Is the best use of his time/energy still the same as when he was starting out?
Clearly the balance is shifting, at least by degrees (I’m *not* saying that the written analysis isn’t still central to the whole edifice, just that the house now has some pretty big add-ons).
Ben is slowly migrating from almost 100% text analysis in his early years (plus some podcasting) to a significant amount of in-depth interviews that *few others can do like he does them* because they’d need an overlap of access AND deep domain knowledge that few possess.
There’s a bunch of journalists and big podcasters out there with access, but their RAM isn’t loaded with the software to do a tech-strategy interview that would satisfy the diehard tech-biz nerds.
Ben also appears to be starting to flex his distribution muscles.
This is why we’re seeing, on top of the podcast interviews, the expansion of the Stratechery cast of characters with the new Sharp China podcast with Bill Bishop from Sinocism, which joins Dithering with John Gruber (🍏) and the recently launched Sharp Tech with Andrew Sharp (🔪).
Will Ben eventually be best-known as a podcaster first, and a writer second? 🤔
How many other shows could be created and instantly find an audience thanks to the Stratechery audience? How many outside shows could be brought into the fold and made more valuable with the increased reach?
As mentioned at the top, there’s always a risk of spreading yourself too thin and diluting what your audience comes to you for in the first place, but by doing it smartly, with the taste to select only worthy add-ons, and by using bundle dynamics/economics to increase the value to subscribers faster than cost increases (if at all), it seems like a good model to attract a wider audience and keep churn low.
🐦🏚🔨👷♂️🪚 Elon Musk’s Twitter Renovations 🚧 🧱
Jessica Lessin ( editor-in-chief of The Information) tweets:
Watching @elonmusk + Co take over Twitter is like watching a business school case study on how to make money on the internet. Amazing that at some level it is so basic. Lesson one: charge your power users what they are willing to pay. Lesson two: Lock up your core customers (advertisers) by going on a listening tour and asking them what works and what doesn't. Make sure to land a few key public commitments. Focus on the head not the long tail. Lesson 3: Cut costs dramatically and do it fast and all at once to get it over with. Elevate a few key people (including some veterans) to promote internal cohesion. Not rocket science. They will succeed because they can do it outside the public shareholder eye.
We’ll have to wait and see if it works, but it’s a good example of what I keep saying:
The important things tend to be simple, they’re just hard to execute and keep in mind consistently.
A lot of businesses go astray because they ‘forget’ what they’re supposed to be doing and instead get distracted by unimportant side-quests, keep copying what competitors do rather than figure out what they should really be doing.
All this made worse by the agent-principal misalignment.
🎧 Companies that win because competitors lack focus 🤔
A discussion of Spotify on Twitter made me think of how, on paper, they’re heavily disadvantaged against giant competitors like Google, Apple, and Amazon.
These can run their music divisions at low margins or negative margins forever if they want to, and it won’t even really register for their overall finances.
Shouldn’t they be able to crush Spotify?
Ah, but there’s something else.
Spotify’s main advantage is that music/audio is their #1 priority.
To their competitors, it's priority #47 (if that).
So the user experience (UX) tends to be the moat.
A bit surprising since I don’t think Spotify’s UX is that great in absolute terms, but it appears better-enough than the competition in relative terms to matter…
Thinking about this made me wonder, who else is in that boat?
Who else has all kinds of disadvantages against strong competitors, except focus 🤔 If you can think of good examples, please let me know in the comments, on the private Discord, or by replying to this email. SPOT 0.00
🛒👕👗👖 AI-generate fast-fashion?
It may get easier to have a really unique ugly sweater soon:
Founded in 2016, CALA is a fashion platform built for designers looking for an accessible way to turn ideas into tangible products. The service is available through both its website and a mobile app. [...]
CALA, an “operating system for fashion” that helps designers sketch, prototype, and produce new products, is the first service to implement OpenAI’s DALL-E API. Its new generative AI tool is live and free to try. (Source)
Facebook Meta Sentiment Check 🎭
These headlines appeared in the same publication, not even a year apart:
🧪🔬 Liberty Labs 🧬 🔭
💡Common Misconception About: Microwave Ovens 💡
Microwave ovens are not tuned to any specific resonance frequency for water molecules in the food, but rather produce a broad spectrum of frequencies, cooking food via dielectric heating of polar molecules, including water. Several absorption peaks for water lie within the microwave range, and while it is true that these peaks are caused by quantization of molecular energy levels corresponding to a single frequency, water absorbs radiation across the entire microwave spectrum.
🇨🇦 ‘Canada commits C$970 million to new nuclear power technology’ ☢️
Canada will provide C$970 million ($708 million) in financing to develop a grid-scale small modular reactor (SMR) [...] designed by GE Hitachi Nuclear Energy [...]
The project, which is being developed by utility Ontario Power Generation (OPG) in Darlington, Ontario, will be the first commercial grid-scale SMR in the Group of Seven wealthy nations (G7) (Source)
Good to see this low-cost funding for an asset that will no doubt pay for itself many many times over during its multi-decade life while providing zero-carbon energy security, but that needs many years of construction during which there’s no cashflow, so if the incentives were set to only look at the short-term, it would be hard to build.
By reducing the upfront financing risk, the government is making it more attractive to build these peerless infrastructure assets (I mean it seriously — what else does what a nuclear power plant can do?).
Human-Safe UV-C light may mean you don’t get sick as often 💡🦠
We’ve been sterilizing things with UV light for a long time (and the sun has been doing it for us outdoors).
Hospitals use lots of UV lights it to sterilize equipment and empty rooms, etc. My parents have a UV light as part of their water-pump system to kill whatever crap is in the water that they get from the nearby lake at their cottage (they don’t drink that water, but it’s still better for it to not have bacteria & viruses).
But unfortunately, 254-nanometer-wavelength light can damage your eyes and skin…
Enter UV-C light:
Now researchers show that a type of shorter-wavelength UVC light, in the far-UV region, that is safe for people can kill 98 percent of microbes floating around in a room. As long as the lights were on, microbe levels remained low even as more were continually sprayed into the room. [...]
This is where UV light with a shorter wavelength of 222 nm comes in. Its higher-energy light is strongly absorbed in the outer layers of human tissue. It does not get past the dead cells on the outer layer of skin or the tears that coat the eyes. [...]
“If the technology had been well developed five years earlier and far-UVC lights had been widely installed, potentially the COVID crisis could’ve been unlike what it has turned out to be,” says David Brenner, director of Columbia University’s Center for Radiological Research, and an author of the paper. “We’ve really lacked the tools to fight COVID. Masks, vaccination, and social distancing all involve people having to make decisions. Far-UVC is a passive device, it’s just up there in the ceiling.” (Source)
The krypton chloride excimer lamps that we currently use are still too expensive and don’t last long enough, but solid-state (think LEDs) UVC lights are being developed, and once they’re widely available, it’s easy to imagine them being installed in all kinds of indoor public spaces (hospitals, stores, train stations, airports) to help reduce the spread of all kinds of pathogens.
🎨 🎭 Liberty Studio 👩🎨 🎥
🧩 A new game to play with kids 🎨🤖
Last night, my 8yo and I played a new game for the first time: We sat in front of my computer and we created images of what he wanted to see using a generative AI (Stable Diffusion, in this case).
Lately, he’s really into the color purple and dragons, so you can see the result above. He’s also a big fan of Zelda: Breath of the Wild…
He loved it, and the fact that it takes a little while for each image to be rendered/created meant that we had plenty of quality time to talk, joke around, plan our next creation, and do push-ups and pull-ups while we waited.
It was quite wholesome, I recommend it if you have kids.