Discover more from Liberty’s Highlights
418: Tech Comp, Twitter, Apple & Microsoft, Prompt Engineers, Taylor Swift's Logistics, David Senra, Oppenheimer, and The Diplomat
"I have a favor to ask you today."
The secret to being productive is to work on the right thing—even if it's at a slow pace. -James Clear
➡️🔔 📩 📺🎙️🎧 I have a favor to ask you today.
Please do it, it will only take two seconds:
Hit “Reply” to this email and send me the name of your favorite Podcast or YouTube channel that you think there’s at least a 50% chance I don’t know about.
By this I mean, something that you love but haven’t seen me mention in the newsletter, and think may be up my alley but isn’t so well-known that there is a chance I don’t know about it (ie. not Joe Rogan).
If you want extra credit on your homework, add a short description of why you like it.
Thank you! 💚 🥃
🟦🌤️ If you’re on Bluesky, you can follow me there:
I’m not yet fully sure what I think of it, but I like to test these things firsthand (more thoughts about Twitter and Bluesky below).
I’m also on Mastodon:
💵 🕰️ 💰 My friend MBI (💎🐕) wrote something that stood out to me while describing his experience at the Berkshire meeting. I thought it was a great example of using people’s mimetic nature in a good way:
One thing I wanted to mention is when people ask question at the meeting, they often mention how long they have been shareholder of BRK. That must create some sort of “cultural” pressure of being long term shareholder of the company for most people who attend the meeting Buffett was so ahead of everyone in building a shareholder base.
If you’re going to feel social pressure, it might as well be in a beneficial direction!
This is probably an underutilized tool.
People know tricks such as making goals public to harness peer pressure and make success more likely, like when it comes to exercising or losing weight.
But that’s basic 101 stuff. If you understand human nature a little, you can help nudge and shape the culture of companies/groups of friends/whatever through similar methods.
Jocko Willink would just call that leadership.
Be the change you want to see, lead by example, from the front, and then make sure to positively reinforce good behaviors. Over time, others will start doing it too — mimetics! — and it will become self-reinforcing in the group.
🛀💭🤰🏻👩🏻🍼🌏 I wonder what the population of Earth would be if humanity had invented safe, convenient, and cheap contraception earlier in our history.
Say in 1800? 1600? 1200? 2000 BCE? 10,000 years ago? 🤔
🥩🍽️ I’m trying to eat better. I eat too much crap in between meals, and wish I had more healthy high-protein quick meals/snacks available (after listening to the Attia-Patrick podcast I link below, you may also want to start eating more protein too to prevent sarcopenia).
One way I plan to achieve that is by using Ethan’s tricks for freezing meat in a way that makes it quick and easy to cook one portion at a time from frozen while retaining flavor and texture:
Everything is a world in itself, and you can keep learning more and more about almost anything.
Because we eat multiple times a day and it has such a big impact on our health and how we feel, on top of the direct enjoyment we get from good food, any cooking/food/nutrition skills & knowledge gained will have huge leverage across your lifetime.
It’s a low-hanging fruit for most people. Chances are you’re like me and are pretty far from the point of diminishing returns when it comes to food.
💚 🥃 🐇 This is the free edition of Liberty’s Highlights with 17,700+ subscribers.
You can get 1-2 extra editions/week full of juicy stuff + access to the private Discord 🗣🗣🗣 community by becoming a paid supporter (it’s quick & easy).
Paid post since last week 🐇:
If you click the link above, you can see the intro for free and there’s a link to get a risk-free 7-day free trial of the paid version. You can only gain from this!
🏦 💰 Liberty Capital 💳 💴
‘Highest median total compensation in tech for 2022’s financial year’ 💰💰💰
Gergely Orosz at The Pragmatic Engineer looked at the 2022 filings for 100 tech companies and compiled the top 10 for the highest median annual total compensation.
Sure, there are some one-off things that happened to make some of these as high as they are in 2022 and they were lower on other years, and with 2023’s layoffs and rebalancing labor market, we’re likely to see these numbers come down. But it’s still pretty incredible for median numbers (they aren’t skewed by a few superstars)!
How are pre-revenue startups supposed to compete for talent without raising ridiculous amounts of VC money if engineers cost that much? That can’t be good for innovation. I wish that bootstrapped startups with ambitious technical roadmaps could execute without having to raise tons of money or issue gigatons of equity…
🐦 🕸️ Twitter: Network effect in reverse? 🍗 😬
We all know it: The network effect is a very powerful thing.
In an ideal network, each new node in the network makes all existing nodes more useful and valuable.
On a telephone network, the more people who have phones, the better your phone gets.
However, not all networks are perfectly balanced, with each node being just as likely to connect with every other node, right?
Twitter has been having a bit of a rough time lately. While I would still bet on them to win against their main competitors (Mastodon, Bluesky), I also want to keep in mind that network effects cut both ways and that in a world of power laws, they don’t have to lose a huge amount of users to feel *real* pain.
If they lose just the top 1-2% of power users that create most of the best content/engagement/halo for the platform, things can unravel rather quickly.
Every super-connected node that leaves the network makes it less valuable/useful for potentially millions of smaller nodes/lurkers.
When it comes to Twitter alternatives, I have a few thoughts:
💭 I agree with John Gruber (🍎) that the community should reclaim the word “tweet” and stop trying to invent stupid-sounding alternatives like “toot”. Tweet should be a generic term like “email” because it’s really a format (everybody knows what a book is, what a memo is, and what a tweet is) and if Twitter had been invented in 1996 instead of 2006, it would probably be an open protocol like email/HTTP/RSS/IRC rather than a private company. The tweet format fills a real need in micro-blogging, and I think Twitter’s erratic stewardship is going to make them lose the monopoly on the format — they’re snatching defeat from the jaws of victory.
💭 While I would still bet on Twitter to be the winner in the space, my #2 choice for most likely winner would be Bluesky over Mastodon at this point. It’s just more user-friendly to non-nerds and is a closer analog to Twitter. So if the bird app unravels, it’s a more convenient drop-in replacement for many users. Never underestimate how big a difference complexity and friction make! And the invite-only roll-out is also creating scarcity/FOMO, which is always a big deal with mainstream audiences (Facebook was invite-only and limited to Ivy League schools during its early days. People seem to care about that).
📖 Decoupling: The Learning Period vs Action 💵
Here’s one that will sound really obvious, but is still worth mentioning:
When it comes to investing, the learning period doesn't have to coincide with the buying period.
Most of the time when something gets really cheap, it doesn't stay that way for very long, especially with very high-quality companies that are normally fairly priced or expensive and so have a huge number of would-be investors waiting in the wings for the stock price to go down so they can buy the dip.
To be ready to strike when the moment comes, it helps to have done your homework well in advance. That’s my reason for studying companies whose business I like even if I think their stock is currently overvalued.
You never know what will happen tomorrow ¯\_(ツ)_/¯
Of course, all the above can be generalized to other fields, it doesn’t just apply to investing. It reminds me of one of the best speeches on TV, from Arnold Rothstein in Boardwalk Empire:
🪦 Requiem for Prompt Engineers 😵🤖
Many people initially thought that “prompt engineering” was going to be a job of the future when generative AI exploded into the zeitgeist. Learning all the incantations required to get the most out of the AI models was surely going to be a prized skill… Right?
This may have been just a mirage. 🏝️
The models are getting better at figuring out what users want at a rate that makes it unlikely that weird command-line-like incantations will be necessary longer term, at least not for most applications. There will always be exceptions, but probably nothing like what was imagined at first.
prompt engineering is an emerging field, but right now it lacks the foundation for long term success. [...] The problem is that more and more prompt engineering will be done by AI systems themselves. I have already seen a bunch of great examples of this in production today. And it’s only going to get better. Just wait for the day when ChatGPT can synthesize your previous conversations and do some auto prompt engineering for you on your queries based on the chi text that it has. All of this is to say, it’s not clear prompt engineering is differentiated long term. This isn’t to say that people who deeply understand how to use these systems aren’t going to be valuable, but I imagine this will be a skill that is learned as part of people’s standard educational path, not some special talent only a few have (like it is today).
In short: If we expect AI to become very good at all kinds of language-based tasks, why would we not expect it to become hyper-proficient at prompting generative AI?
And the chatbot format is particularly adapted to a back-and-forth iterative collaboration that refines intentions and thus doesn’t require a 0-shot perfect prompt from the start.
🎤🎶 Taylor Swift’s Tour: Logistics and Financials 🛩️ 🚚🚚
I don’t consider myself a fan of Taylor Swift, though I have heard a few of her albums a handful of times each and respect her talent as a musician and as a businesswoman.
I saw a documentary a few years ago, I believe it was Miss Americana, and it was a nice behind-the-scenes look at the megastar touring life.
This thread by Craig Fuller about her current insanely big tour was very cool. Here are some highlights:
The logistics and transport costs alone are more than $30M for the tour.
A trucking fleet of 90 at least trucks will stay with the tour the entire time and cost as much as a half million dollars a week [...]
Nothing is left to chance. They have duplicates and backups of everything. [...]
It’s like the saying in the military: Two is one and one is none.
They use multiple stages and crews in different cities to keep the tour rolling. One city may have its concert that week, while another is being built out at the same time. [...]
Forbes estimates that she will personally gross $620M and net over $480M after all her expenses have been paid. This will double her net worth and crown her Era Tour as the most profitable concert in history.
Half a billion dollars from one tour! And she’s not going on tour buses, she’s flying from city to city on a Falcon 900 jet.
🍎📲💰 Apple's revenue from the iPhone alone is more than Microsoft's entire revenue 📊
☕️ Interview: David Senra 🔥
Listening to David (📚🎙️) is always like a few shots of espresso, so if you need a bit of a pick-me-up, check out his interview on the My First Million podcast:
💾 ⏳The durability of software businesses
Whenever someone questions the durability of software businesses I remind them that Teradata and Dropbox are still alive and generating hundreds of millions of FCF/yr
He’s got a point.
While many software businesses never earn their valuation in the market or fail to reach dominance and grow gigantic the way they are expected, in absolute terms many of these are very profitable and sticky businesses.
They only fail at the expectations game.
🧪🔬 Liberty Labs 🧬 🔭
‘We eradicated smallpox for less than it costs to build one mile of subway today in NYC’ 🦠 🚇
The title is a line by Jason Crawford.
He measures a few other things in NYC subway miles-equivalent:
Panama Canal: ~4 NYC subway miles
Manhattan Project: 10 miles
GPS: 10 miles
Human genome: less than 3 miles
Operation Warp Speed: 5 miles
It’s a great intuition pump, worth pondering in both directions: ie. why don’t we do more of these high-impact projects, and why has it gotten so long and expensive to build anything in the West?
It’s also really interesting to see how expensive and how lenghty these massive projects were.
$500bn in today’s dollars for the National Highway System in the US! 💰💰💰💰💰🛣️
$280bn for the Apollo program! $255bn for the ISS! 💰💰💰🚀
🥼🩺👩🔬🧬 Interview: Peter Attia & Rhonda Patrick 🔥🔥🔥
This was a great podcast and I highly recommend it if you are human and have a body.
They touch on so many important topics, basically covering all the main causes of death in wealthy countries *and* ways to prevent/delay them.
Alzheimer’s disease, cardiovascular diseases, cancer… But the discussion is at a very applied level, with recommendations (sauna, omega 3, fasting or not? how much protein in diet? sleep quality? etc).
Let’s just say that it really made me want to exercise more, so if you’re looking for a bit of motivation on that front, this is a good one.
I’ll be listening to it more than once. It’s one of those reference episodes IMO.
And to be clear, these are the things you should be worrying about, not what you see on the news:
⚛️ Nuclear Waste: Important Context ☢️
But what about the spent nuclear fuel — isn’t it radioactive for hundreds of thousands of years? The way radiation works, the waste products that are the most radioactive are the shortest-lived, and those that last a long time are far less dangerous. About 40 years after the fuel becomes waste, the heat and radioactivity of the pellets have fallen by over 99 percent. After around 500 years, the waste would have to be broken down and inhaled or ingested to cause significant harm.
Compare this to other hazardous industrial materials we store in less secure ways that don’t become less toxic over time.
Take ammonia: It is highly toxic, corrosive, explosive and prone to leaking. Hundreds of ammonia-related injuries and even some fatalities have been reported since 2010, and we continue to produce and transport millions of tons of it annually by pipelines, ships and trains for fertilizer and other uses. [...]
The waste should really be a chief selling point for nuclear energy, particularly for those who care about the environment: There’s not very much of it, it’s easily contained, it becomes safer with time and it can be recycled.
This last part refers to how “spent” nuclear fuel still contains something like 95%+ of its energy, so future reactors using existing designs (breeders, etc) will be able to use this “waste” to generate clean power for many more decades into the future.
🎨 🎭 Liberty Studio 👩🎨 🎥
💣 Christopher Nolan’s Oppenheimer 🎬
I hope Richard Feynman gets a good amount of screen time and they do a good job with his character!
The other two films I’m looking forward to this year (along with 👆) are Dune: Part 2 and Asteroid City by Wes Anderson. What about you?
📺 The Diplomat Season 1: final thoughts (mild spoilers) 👍
My wife and I have finished watching the first season of this show on Netflix.
Overall, I have to say that I really enjoyed it. The final few episodes felt a bit rushed at times, and I kind of wish they had a longer season (maybe 10-11 episodes?), but it was still a very good first season.
The ending was the most Hollywood thing about it, but hey, I get it, they needed a twist, they needed to increase the stakes for the main characters and leave us hanging, wanting more.
I can think of a few ways to handle the situation and the “bad guys”, and I hope that they are able to steer things back quickly into the show’s sweet spot. It would be a shame if it turned into The Bourne Ambassador. 😬
I hope Keri Russell wins Emmys for this performance 🏆 She truly was excellent and shone in almost every scene. Rufus Sewell was also very good as Hal, especially since he has to walk a fine line the whole time to keep us guessing and cycling between hating and loving him.
The supporting cast was also very strong. Ali Ahn as Eidra and Ato Essandoh as Stuart had great chemistry, and almost everything out of Eidra’s mouth was gold.
Standing back, the writing was good, the cinematography was good, the sets and locations were beautiful, the acting was good. It’s not perfect, there are some flaws, but there’s plenty to like and it worked great as a kind of procedural thriller. I recommend it 👍