Discover more from Liberty’s Highlights
442: LK-99, Nvidia, Peter Thiel, Twitter, Waymo, Hollywood vs YouTube, AMD + India, Wind & Solar, and Roman Industrial Revolution
"psychic vampires that just want to drain your energy"
Status games are fun to analyze and horrible to play.
👨🔬🧲 With the caveat that prediction markets may not be all that effective with one-off unprecedented discoveries, it’s still nice to see the odds of the LK-99 room-temperature superconductor replicating moving up (and even flipping for a moment).
I can't wait to have more certainty one way or the other on this...
Room temp superconductors would be such a world-changing thing that there are currently two very different timelines superimposed in my mind when I imagine what’s ahead.
Like an uncertain quantum state, they will soon collapse into one or the other as we learn more, but it’s a bit nerve-wracking until then 😬
I’ve had my heart broken too many times, so I kind of still expect a rug pull at the last minute, but it’s looking possible!
To recap, preprints appeared last week making the remarkable claim of a well-above-room-temperature superconducting material at ambient pressure, dubbed LK-99. This is one of the most sought-after goals in all of materials science and condensed matter physics, something that until now has only been found in (numerous!) science fiction stories. [...]
Well, extraordinary claims need extraordinary proof, and where things like this usually fall apart is difficulty with replication. [...]
As of [the] morning [of August 1st], there are (as yet not really verified) reports of replication from the Huazhong University of Science and Technology in China. At least, a video has been posted showed what could be a sample of LK-99 levitating over a magnet due to the Meissner effect, and in different orientations relative to the magnet itself. [...]
two other new preprints interest me greatly. One is from a team at the Shenyang National Laboratory for Materials Science, and the other is from Sinéad Griffin at Lawrence Berkeley. Both start from the reported X-ray structural data of LK-99 and look at its predicted behavior via density functional theory (DFT) calculations. And they come to very similar conclusions: it could work. [...]
I am guardedly optimistic at this point. The Shenyang and Lawrence Berkeley calculations are very positive developments, and take this well out of the cold-fusion "we can offer no explanation" territory. Not that there's anything wrong with new physics (!), but it sets a much, much higher bar if you have to invoke something in that range. I await more replication data, and with more than just social media videos backing them up. This is by far the most believable shot at room-temperature-and-pressure superconductivity the world has seen so far, and the coming days and weeks are going to be extremely damned interesting.
🕵️♂️🤔🤹🏻♂️ Paul Millerd has a tweet (no, definitely not an 𝕏) that made me think about applied curiosity:
Most adults don’t really learn that many new things and I think a lot of it is that we don’t know how to get from curiosity to actually starting to learn something. It’s much more about hacking your state of mind and level of excitement than anything that resembles school
It made me think of a French expression that goes: "L'appétit vient en mangeant"
Literally: "One becomes hungrier as one eats", or “the more you eat, the hungrier you get”.
Curiosity is kind of like that.
You just have to get started and it'll generate its own momentum eventually. Keep going until you get to that point, and then you probably won’t be able to stop even if you wanted to!
It’s a skill.
Over Time, you get better at learning new things, experimenting, keeping an open mind and not defaulting to “not for me” too quickly, at giving something new a real try (most people stop too quickly or pre-judge how much they like something based on not much — nobody would like coffee or beer if they only ever had one sip).
The implications of this are profound:
If true, it means that the hardest it’ll ever be is at the beginning.
It’s an encouraging thought! Get started now and get good at it so you can get to the more flow-y, more enjoyable part!
💚 🥃 🐇 In case you missed them, I posted my 600+ word review/thoughts on the film Oppenheimer in Edition #441 and my review/thoughts on Miyazaki’s Spirited Away in Edition #439. These editions are not just reviews, they were filled with other goodies about investing, business, science, and technology.
These are for supporters, but there’s a 7-day free trial if you want to explore.
🏦 💰 Business & Investing 💳 💴
Peter Thiel used to pay people an extra $1,000/month to live within a half-mile of the office 👔🗄️💼🚶🏡
This is from the book Paper Belt on Fire by Michael Gibson. I haven’t read it, but JAKE shared the excerpt above.
It’s an interesting idea — it could be interpreted purely cynically, like, I only want people who will live at the office and work all the time and be easy to reach outside work hours. I’m sure that’s part of it.
But in my experience, short commutes are such a quality-of-life improvement that you probably end up with happier and healthier employees who aren’t spending hours each day stuck in traffic.
But we’re also not talking about working at IBM or HP, right?
Startups can be extremely intense. It’s a trade-off that many people willingly make to have a lottery ticket, or to work on more interesting things, or because they like small teams where a single person can have a big impact in shaping a project. I think that removing long commutes from the equation makes more of a difference in that context.
If I try to imagine my young self as a participant in those stories, I think reclaiming commute times would make a big difference. My own way to do that has been to work remotely for the past 20 years, but that’s not always possible.
🚂💰💰💰 Twitter stock grant gravy train is ending…
You’ll recall that when Musk bought Twitter, he paid a rich premium: $54.20 a share. Some of the 1,350 or so X employees are still receiving regular stock grants at that price, and the occasional windfalls have been enough to keep them at their jobs over the past year of tumult.
Tomorrow, those employees will receive their final grant of $54.20 stock. And the question is how many of them will stay much longer.
🚚🤖 RIP Waymo’s autonomous trucking program 🪦
Given the tremendous momentum and substantial commercial opportunity we’re seeing on the ride-hailing front, we’ve made the decision to focus our efforts and investment on ride-hailing. [...]
we’ll push back the timeline on our commercial and operational efforts on trucking, as well as most of our technical development on that business unit. We’ll continue our collaboration with our strategic partner, Daimler Truck North America (DTNA), to advance technical development of an autonomous truck platform.
This is the positive spin on it: So much opportunity in ride-hailing, we want to focus on that!
While Waymo insists it still intends to apply its autonomous vehicle system to trucking, it didn’t provide further details on when it might renew the program or launch a commercial business.
The vast majority of employees on Waymo’s trucking team have taken other roles within the company. A few number of individuals will be affected by the change but will be helping with the wind down of the program. Waymo did not provide further details on exactly how many people may be impacted.
Too bad. Trucking is an area where autonomy could make a big difference.
There are labor shortages and the median trucker is getting old. More efficiency could be brought to transportation with trucks that could draft in convoys, drive safely for more hours per day or at night when roads are empty and there’s no traffic, etc.
🎬🔐 Hollywood’s Pain is YouTube’s Gain 📺
The Hollywood labor strikes have proven a boon for YouTube, with Google's video-streaming service luring television ad buyers who are worried about committing to traditional TV ads while actors and writers remain on the picket lines
Elon Musk to utility sector: Higher demand is coming, plan for it 🔌⚡️🏭 📈
Tesla CEO Elon Musk told attendees at a summit held by Pacific Gas & Electric that “whatever your demand predictions are for electricity, I suspect they are too low… I recommend anticipating much higher demand.” [...]
Global energy consumption today is roughly a third electricity, a third for transport and a third for heating, he said.
“So even if you assume that electricity demand is static, in order to transition to a sustainable energy future where everything is electric and sustainably electric, we need a tripling of electrical output,” he said.
None of this happens overnight, but I suspect it will be very non-linear. Because of the long lead times associated with large power plants, new transmission capacity, and dealing with regulatory friction, there’s no time to waste.
If we wait until the demand has materialized and is staring us in the face, it’ll be too late.
☁️👑☁️ Nvidia plays kingmaker to increase cloud competition + Sniff out startup investments 🤔
Not long ago, I wrote about how Nivida sending a lot of supply of its scarce H100 data-center GPUs to smaller cloud players to help them better compete with the hyperscalers.
Over the past few months, Nvidia has supported several startups—including CoreWeave, Lambda Labs and Crusoe Energy—that aim to compete with larger cloud providers. Nvidia has been ensuring that these upstarts get early shipments of its most advanced AI chip, the H100
From Jensen’s point of view, it makes a lot of sense, because the more customers he has, the less pushback he’ll get on his margins.
He has a frenemy relationship with the hyperscalers anyway.
They buy a lot of his chips, but they’re also working on their own in-house AI accelerators (Google’s TPU, AWS’ Inferentia, etc) and software frameworks to abstract away the hardware layer and get around CUDA (Meta has been doing a lot of work on that).
Now Nvidia is taking that to the next level:
Nvidia began asking the smaller cloud providers about their customers in the past few months, according to one of the people with direct knowledge. [...]
Knowing which startups are requesting large numbers of GPUs also gives Nvidia more insight into the market for AI applications and could prompt the company to strike more-direct relationships with those customers
There may also be a speed angle:
Nvidia seems to be growing increasingly impatient with the larger cloud providers, who have been slow to make the company’s H100 chips available to customers. Google Cloud, Azure and Oracle offer H100s in a private preview, and AWS just made the H100s available to all customers on Wednesday.
“We’re really frustrated with the large cloud infrastructures…not giving people access,” Robert Ober, Nvidia’s chief architect for hyperscale data centers
🇮🇳 AMD to open a large R&D center in Bangalore, India 🐜
today announced plans for continued growth in India through an approximate $400M investment over the next five years. The planned investment includes a new AMD campus in Bangalore, Karnataka that will serve as the company’s largest design center, as well as the addition of approximately 3,000 new engineering roles by the end of 2028.
AMD has around 6,500 total employees in India today, so this is a significant investment.
As many chipmakers try to diversify their footprints globally, it’ll be interesting to watch what happens in India.
h/t Dan Nystedt
⚖️ Brian Chesky's advice on how to deal with regulators 👂🏻
This is from a talk at Stanford's Graduate School of Business:
Brian Chesky: Oh, man. Yeah. There were some things that were initially out of my depth. Like they did not teach you at RISD how to work with government regulators. So I always grew up thinking, if people don't like you, you should avoid them. That seems kind of intuitive, right? Or you should fight them and disagree with them and they're your opponent.
And then in 2011, I hired my first executive. Her name is Belinda Johnson. She was our chief legal officer, eventually became our COO. And she told me something that was counterintuitive to me. She said, if people don't like you, you should meet with them. And I said, really? Why would I do that? Wouldn't it go horribly wrong? And she said, no, because it's hard for people to hate you up close.
And I think that's a really important lesson. So I said, okay, well, who doesn't like me? And it turned out a lot of people. And so instead of having a tour where I went on tour to meet everyone to like me, I told our team I want to meet everyone who doesn't like me. And it turned out they kept me busy for a long time. But I would meet with these government officials and sometimes the meetings would start. A little hostile at first, but I had a rule that would always listen first and seek to understand them. And I think my dad used to say, like, 90% of life is just showing up. It's actually not true in tech.
But when it comes to showing respect, 90% of life sometimes is showing someone respect and listening to them.
This applies more broadly than just regulators.
The trick is to be able to differentiate between critics who are a waste of time and those who may have something useful to teach you.
Some are psychic vampires that just want to drain your energy and should be avoided (plenty of garlic!), but others have genuine and thoughtful disagreements or useful alternative perspectives that you should be aware of and incorporate into your thinking.
🧪🔬 Science & Technology 🧬 🔭
☀️ U.S. Wind and Solar capacity installed per quarter 🌬️
New solar energy installations in the first quarter of 2023 fell 34% from the year-ago period to 1,587 MW. The solar energy sector is exiting a tumultuous period as the amount of capacity installed in 2022 fell 22% from the prior year amid supply chain bottlenecks, trade restrictions and phasing out renewable energy tax credits
The amount of US wind power capacity installed in Q1'23 dropped 47% from the year-ago period, but the five-year pipeline for wind power projects increased by 5% from the fourth quarter of 2022. In all, developers energized 1,472 MW of wind power capacity in Q1'23, compared to 2,785 MW of capacity connected to US power grids in Q4'22
One thing to remember when you look at these numbers is that nameplate capacity doesn’t tell you how much actual electricity you produce.
Every power source has its own characteristics, and in the case of wind and solar, the capacity factor will vary depending on the project, but tends to be somewhere between 20% and 40% (so a 1GW project will produce between 200-400MW on average).
🇺🇸 Americans are moving around! Which areas are growing vs shrinking? 🏡🏠🏘️ 🏚️🏚️
Note that because U.S. fertility is not particularly high anywhere (even Utah), most of this growth is going to be from people moving around, rather than from births and deaths [...]
in general, Americans are still moving west, especially to the Rockies and beyond, and to Texas and the Southwest. The exception is the California coast, which people appear to be fleeing despite continued robust immigration from Asia and Latin America. The obvious reason is housing; California cities just don’t build much of it. Meanwhile, Americans generally continue to move from the northeast to the southeast, and from rural areas to big cities.
If you couldn’t live where you currently live (in the US or outside of it) and had to move somewhere in the US, where would you go?
Is it a big blue dot? Are you a contrarian who would move to a big red dot?
🚂🏛️ Why no Roman Industrial Revolution?
Friend-of-the-show Jason Crawford has a great post on this:
How are you going to industrialize when you don’t have cast iron to build machines out of, or basic mechanical linkages to use in them? How could a society increase labor productivity through automation when it hasn’t even approached the frontier of what is possible with simple wooden tools? Why even focus on improving labor productivity in manufacturing when productivity is still very low in agriculture, which is more fundamental? Why should it exploit coal when it has barely begun to exploit wind, water, and animal power? How are engineers to do experiments and calculations without any concept of the experimental method, and without anything close to the mathematical tools that are available today to any fifth-grader? And if anything was discovered or invented, how could the news spread widely when most information was hand-written on parchment?
All of the flywheels of progress—surplus wealth, materials and manufacturing ability, scientific knowledge and methods, large markets, communication networks, financial institutions, corporate and IP law—were turning very slowly. There is not a single, narrow path to industrialization, but you have to get there through some path, and ancient Rome was simply nowhere close.
🎨 🎭 Arts & History 👩🎨 🎥
🌺🌸 Behind the Scenes of Nirvana Unplugged 🌸🌺
After decades of seeing this show only from certain angles, it’s really strange — in a good way — to see this footage of them getting ready for the performance and generally having a good time.
🎮 Interview: John Romero (Doom and Quake co-creator)
I enjoyed this interview by Tim Ferriss:
He mentions his playthrough of MyHouse.WAD, which I have linked in Edition #435 — that got a smile out of me!
His discussion about the early days of id Software and how they made games at the time is fascinating. It helps that John has an incredibly good recall of it all.
Some years ago, I read the great book ‘Masters of Doom’, and while I was always more of a Carmack nerd, partly because at the time Romero seemed to be in a high-hubris phase (Ion Storm/Daikatana 😬), it’s nice to see that he’s come very far since then. He seems like a good dude.