369: Amazon Ads vs Facebook, EV Batteries, Credit Card Fees, TSMC, Shoplifting, Mauboussin, Japanese Knives, ChatGPT, and The Last of Us

"Making something look easier than it is is part of mastery."

It is not the mountain we conquer, but ourselves. —Sir Edmund Hillary (1919-2008), first man to climb Mt. Everest

🎌🔪👨🏻‍🍳 On Xmas eve, after my wife and kids were asleep, I fell into a kitchen knife rabbit hole. I have a decent collection of folding knives — a hobby from before my kids were born — but I hadn’t really looked into kitchen knives yet. Until now.

I’m specifically interested in Japanese knives.

The main benefits vs European-style knives seem to be:

• They’re extremely sharp because they tend to be made from harder steel (often carbon steel, which can rust, so you have to wipe them dry after use) and have thinner blades. However, this makes them more prone to chipping.

• They’re very light due to the thin blade and light wooden handles.

• Many are still hand-made by blacksmiths in Japan. They have a distinctive look that may or may not be for you, but I find them very aesthetically pleasing. There are all kinds of finishes, from very rough to shiny multi-layered Damascus steel.

Don’t be surprised if there’s some knife-related content in future editions — hopefully it piques your curiosity and improves your knife game!

I’d like a nice Gyuto. What’s the difference with a Western chef’s knife?

The way I see it, if having a nice knife makes cooking more enjoyable, it creates a virtuous cycle where you do more of it, level up your skills, and then benefit from better home-cooked food forever after.

🕰️🔭🌍 There’s a big difference between believing we can know the future and believing that we can know some of the risks we’re facing.

Without having any idea what is going to happen specifically, we can still try to figure out which risks are worth mitigating to increase the odds of a good future, whatever it is.

Anything that would have a very high impact — all the way to existential risks — should be taken seriously, even if probabilities are relatively low.

In fact, these should be prioritized over very high-probability risks that have very low impacts especially because the human mind tends to focus on what’s most obvious AND easy to fix because we love the feeling of progress and accomplishment, even if it is minimal, over the feeling of banging our heads against the wall and uncertainty about having any impact at all.

ie. All the energy and mindspace that went into banning plastic straws could’ve made orders and orders of magnitude more impact if it was directed in a useful direction.

🤔 Morgan Housel asked in a tweet:

what's something that's harder than it looks?

My answer: Anything done right.

Making something look easier than it is is part of mastery.

🐦 Very thoughtful piece by Bhogal Gurwinder about the ‘Dramageddon’ on Twitter and what to do about it. Some highlights:

The nature of social media as a realm of pure appearance, in which one can easily deceive others, makes it highly attractive to narcissists.

Narcissism is pathological self-centeredness, and it’s characterized by a strong sense of entitlement, a need for attention and social validation, and a lack of concern for the needs of others. In order to get maximum attention, narcissists will often engage in histrionics (theatrical behavior), catastrophizing (exaggerating) threats against them. They are therefore frequently surrounded by an aura of drama. [...]

To understand how a minority of people can drive the majority of online toxicity, we must consider social media’s second visual filter. Not only does it attract narcissists, but it also amplifies them. [...]

The high engagement that online belligerence receives has a double effect: it means that toxic posts are more widely circulated, but it also means that people who act toxically are rewarded. Rewards are incentives [...]

None of this would be problem if it were inconsequential, but the consequence of believing that people are extremer than they are is that it makes you extreme.

Worth keeping in mind as we navigate social media.

---

🧩 A Word From Our Sponsor: Heyday 🧠

Do you have 100+ browser tabs open right now? 😬

Give your memory a boost with Heyday, the research tool that automatically saves the content you view, and resurfaces it within your existing workflows. 👩‍💻

It’s like cheat codes for your memory. 😲💡

🧩 Give your memory a boost today 🧠

---

🏦 💰 Liberty Capital 💳 💴

🛒 Amazon Advertising Advantages over Facebook Meta

In Practise:

A: One thing that, for me, Amazon and no one else had, in the media landscape, was such great data. [...] when I thought about the data that Facebook owned, I thought the Amazon data was much more valuable and much more granular, reliable and accurate than the Facebook data. [...]

Q: How do you compare the two today? Facebook and Instagram's data versus Amazon, what are the major differences?

A: The difference is that Amazon has older, purchased data, which is something that Facebook doesn't have. Facebook, the data that they have is declarative data. It's the data that you share when you open your profile on Facebook or Instagram. Then what they know is the accounts you follow, how much you interact with the ads and things like that; your centers of interest. But they don't know exactly what you buy, the products you're looking at and so on. It's way more limited than the Amazon first-party data.

They go on to mention that even for advertisers who *don’t even sell their products on Amazon*, this 1st party data can be incredibly valuable and can be accessed through Amazon’s Demand Side Platform (DSP).

Advertisers can buy ads on Amazon’s properties (ie. Fire TV) that use this 1P targetting data to drive traffic to non-Amazon destinations. AMZN 0.00%↑

📱 Tech’s Economies of Scale + low marginal unit cost 🔨

Maybe I’m easily impressed, but every once in a while, I can’t help but think about the disparity between incredibly sophisticated tech products and much less complex other products/labor.

I can buy an iPhone for around a thousand bucks. It’s the product of thousands and thousands of highly skilled people around the world working together at the cutting edge.

Think of the whole supply chain. 🌐

ASML engineers in the Netherlands are doing black magic stuff to make sure that TSMC in Taiwan can make chips with features measured in single nanometers…

Various companies in South Korea and Japan are building complex parts like memory chips, high-resolution high-dynamic range OLED screens, antennas that can transmit gigabits/sec over the air using incredibly complex physics tricks, multiple cameras that all fit in an area the size of a postage stamp yet take better photos than cameras that weighted multiple pounds not long ago.

There are armies of people assembling all these parts together along the way. What an incredible coordination problem to make all this synch up! One tiny part is missing and you can’t make the final product. 🔩

Then you have the designers, engineers, and UX people creating the thing in the first place, going through iterations and prototyping for the physical objects, the custom silicon, and armies of software engineers who write, test, and maintain millions of lines of code for the operating system, the apps that come with the thing, and then manage a platform of 3rd party apps that use frameworks and APIs that also need to be written and maintained.

Additionally, security people need to make sure that all this is hardened against attacks from script kiddies, criminals, and nation states.

All these people can work for me — get parts manufactured, shipped around the world, assembled, and then shipped to me — for about a thousand bucks.

But if I want to hire one guy to go buy some wooden planks and build a small shed in my backyard, it’ll cost multiple thousands of dollars. 🪚

¯\_(ツ)_/¯

🔋🔋🔋🪫 EV battery manufacturing may stop being viable in Europe 🇪🇺

VW Brand Chief Executive Officer Thomas Schaefer wrote:

The fact is: On the international stage, Germany and the European Union are rapidly losing their attractiveness and competitiveness. [...]

Europe lacks price competitiveness in many areas. When it comes to the cost of electricity and gas, in particular, we are losing more and more ground. Unless we manage to reduce energy prices in Germany and Europe quickly and reliably, investments in energy-intensive production or new battery cell factories in Germany and the EU will be practically unviable. The value creation in this area will take place elsewhere.

😬

💳 In Defense of Credit Card Fees 💸

Friend-of-the-show and OG Extra-Deluxe supporter David Kim (💚💚💚💚💚 🥃 🎩) took a deep dive into payments. Here are some highlight:

“we find significant evidence of banks offsetting Durbin losses by raising other account fees. The share of free basic checking accounts (accounts with a $0 monthly minimum for all customers, regardless of account balance) decreases from 60 percent to 20 percent as a result of Durbin.” [...]

It’s fair to say that interchange is a costly fee for card-accepting merchants. But cutting it appears to impose net costs on consumers in the form of higher banking fees and lower rewards. So in a two-sided card network, interchange reduction basically amounts to a zero-sum wealth transfer from consumers to merchants. [...]

Consider this screechy article from The Week, which declares: “There is no possible moral or economic justification for these fees. Credit card companies and their bank allies are just abusing their market power to soak the rest of society for easy fat profits” and “the immense profits of Visa, Mastercard, and the card-issuing banks are sheer parasitism”.

No possible moral or economic justification? Sheer parasitism? That’s a little intense. Thanks to the card networks, merchants realize more sales volumes than they otherwise would and consumers enjoy rewards, convenience, and global acceptance. Because of continuous investments in fraud management and the collateral requirements that card networks set for member banks, you can hand your Visa or Mastercard to any one of 170mn merchants around the world, complete a transaction within seconds, and know that if you’re ripped off you will be made whole. The merchant, without knowing anything about you, can rest assured that money from your bank will reach theirs. For enabling all this, the card networks take just ~0.2% of the transaction value. That’s not a scam; it’s a marvel.

It’s very easy for most people to attack big, successful corporations for their high-profit margins.

What *really* matters is not how much value they are capturing, but how much value they are *creating* and is that much bigger than what they’re capturing..?

V 0.00%↑ MA 0.00%↑

🇹🇼 TSMC Celebrating start of 3mn Mass Production 🐜

[TSMC] will hold a ceremony on Dec. 29 to celebrate the start of mass production using its 3 nanometer process and plans for further expansion of the technology in southern Taiwan. [...]

TSMC is also developing the more sophisticated 2nm process and will build a 2nm fab in Hsinchu with mass production scheduled to begin in 2025.

The 3nm process uses 16nm FinField-effect-transistor (FinFET) technology, which is a 3D transistor... Compared with the 5nm process, which started mass production last year, it will raise speeds by 10-15 percent and be 25-30 percent more efficient in energy use.

Source. TSM 0.00%↑

✉️ MBI 2022 Annual Letter — Looking back on mistakes 🗓️

I really enjoyed this very introspective end-of-year retrospective by my friend (💎🐕):

• MBI Deep Dives: 2022 Annual Letter

I think we all should spend more time thinking about our mistakes, blind spots, and flawed thinking. The less we think we need to do so, the higher the likelihood that we *really* should.

📈 Bull Market in Shoplifting 🛍️🫲 👀

The National Retail Federation estimates that shrink—an industry term for loss in inventory—amounted to roughly 1.4% of retail revenue in 2021, or roughly $94.5 billion. Most of that shrink is caused by theft. [...]

Although shrink is a perennial problem in retail, it really took off when the pandemic hit. In the five years leading up to 2019, retail shrink grew at a compound annual growth rate of roughly 7%, according to data from the NRF. In 2020, it jumped 47%, and rose another 4% on top of that huge jump in 2021.

Here’s what this looks like:

🐻📉 California’s population shrinks for a third straight year as high costs stress households 🏃‍♂️🏚️🤰🏻🏡💸

Births outnumber deaths in California, and yet the U.S. Census Bureau says the population shrank again as more than 300,000 people moved out of the Golden State.

This is the 3rd year of decline. But 2020 was the *first* year in the state’s history when population declined 😬

Between July 2021 and July 2022, the number of California residents dropped from 39,142,991 to 39,029,342 — a loss of about 114,000 people, the new data shows.

There were more births than deaths in the state, and 125,715 immigrants made their homes here in that one-year span.

Most of the population decline was explained by the 343,230 people who moved to other U.S. states. (Source)

One of the factors blamed for this is the lack of affordable housing, which is hard to fix in a state that makes it hard and expensive to build new housing and increase density…

✂️ Tech Layoffs 🚀

Friend-of-the-show and Extra-Deluxe supporter (💚💚💚💚💚 🥃) Byrne Hobart writes:

over the course of 2021, [Layoffs.fyi] tracked a total of 15,000 or so tech layoffs. This year, the number is 151,648. To put this in perspective, every quarter starting with Q2 of this year had more tech layoffs than all of 2021

One more example of how most changes in life are non-linear because in complex adaptive systems, feedback loops tend to create cascades and accelerating moves. Things can happen rather quickly especially when they reverse direction.

🇬🇧 Energy crisis in the U.K. 🕯️

The UK is encouraging consumers to cut their energy use as part of a campaign aimed at scaling back demand for power and natural gas in the face of soaring bills this winter. [...] It includes advice such as... taking measures to reduce heat leaking through windows and doors. [...]

In Britain, energy bills for many consumers have more than doubled in the past year amid a historic energy crisis for the region, exacerbated by Russia curtailing gas supplies to Europe.

A separate program run by National Grid, which encourages UK households to curb electricity demand at peak times (Source)

One of the richest nations in the world has to cross its fingers and hope that we don’t have a particularly cold winter…

The problem is not that a bunch of things went wrong — things always go wrong, eventually — the problem is getting to this vulnerable point over decades of bad decisions when it comes to energy policy.

🎙️ Michael Mauboussin returns to ILTB (Podcast)

Mauboussin is always thoughtful and thought-provoking. I enjoyed this chat with Patrick O’Shaughnessy (☘️):

• 🎧 Michael Mauboussin - Sharpening Investor & Executive Toolkits

🌎 The 15 largest employers in the world (5 are militaries)

Amazon is right behind China’s People's Liberation Army 😯

🇺🇸 Why the U.S. is so stuck 🏗️✋🛑

When you add the anti-science bias of the Republican Party to the anti-build skepticism of liberal urbanites and the environmentalist left, the U.S. seems to have accidentally assembled a kind of bipartisan coalition against some of the most important drivers of human progress. To correct this, we need more than improvements in our laws and rules; we need a new culture of progress.

Source.

---

🧪🔬 Liberty Labs 🧬 🔭

🔥 Why ChatGPT is a Big Deal and How to “Prompt” it to get Better Results + Pseudo OS and Software Guitar Pedal 🤖

You’re probably getting a bit tired of ChatGPT by now, but it’s worth watching this one.

The example that he gives where ChatGPT fixes a code error from the code that it itself just wrote is 🤯

Or the pseudo-OS thing. That’s kind of nuts. He could actually edit a pseudo-python file in the pseudo-Nano editor and then pseudo-run the code 🤯🤯🤯

The first comment on Youtube is:

ChatGPT helped me solve a university master's/PhD level assignment in advanced probabilistic machine learning. It gave me a valuable insight into the structure of the problem and what kind of algorithm could be used to solve it, and as soon as I saw the suggestion I knew it was right. That kinda blew me away, we are definitely on the verge of an intelligence explosion in AI that will revolutionize pretty much every aspect of society. This will be an interesting decade.

Or how about this one: Someone used ChatGPT to code software to replicate the sound of an electric guitar effect pedal, just by describing it and asking for it.

It’s not perfect, but that it can even do something like this without having been specifically trained at all to do this kind of work shows that its intelligence/capabilities are fairly generalized, and they’ll only get more so over time (GPT-4, GPT-5..).

⚛️ 🇺🇸 NRC’s “Simplified” New Regulations are More Complex and Burdensome Than Existing Regulations 🤦‍♀️

A depressing read from the Breakthrough Institute:

Three years ago, Congress passed the Nuclear Energy Innovation and Modernization Act, directing the Nuclear Regulatory Commission (NRC) to develop a modernized licensing framework that could better accommodate innovation and support the commercialization of a new generation of advanced nuclear reactors.

So far so good. The best of intentions…

Congress explicitly directed the NRC to create a risk-informed, performance-based licensing framework flexible enough to accommodate a broad range of future advanced reactor designs, in contrast to the prescriptive rules, purpose-built to license large light water reactors, that have governed nuclear licensing since the NRC’s formation in the 1970s. Under the old rules, the NRC has not, to date, ever licensed a new commercial reactor design that was subsequently entered into operation.

Read that last sentence. How many never even tried because it didn’t seem worth it?

On September 30, 2022, commission staff unveiled a 1252-page draft proposal, 22 months in the making. [...]

NRC staff claim they have met the mandate given to them by Congress. But the sheer length of the proposed regulations alone demonstrates this is not the case.

The draft framework is twice as long as either of the legacy, prescriptive licensing frameworks, Part 50 and 52, that it is intended to supplant. That is because the staff largely cut and pasted the old rules into the new framework, then added further burdensome regulations, including qualitative health objectives that cannot be complied with and expanded requirements for the notorious “As Low As Reasonably Achievable” radiation standard, a further invitation to endlessly ratchet regulatory requirements.

These latter two standards have been added by NRC staff despite longstanding and clear direction from the Nuclear Regulatory Commission not to use either standard in the way that staff proposes to use them.

An early draft of these new rules was shared with advanced reactor developers, and 85% of them said they’d rather use the old rules because the new framework “appeared to be even more complicated and burdensome than the existing rules”

Ugh.

the NRC staff then included a poison pill in AERI, which requires that the risk analysis assume that a maximum accident occurs every year for the lifetime of the reactor, an assumption that is physically impossible (there is no plausible world in which a reactor could have a maximum accident, rebuild and restart within a year, and then continue to have maximum accidents, rebuild, and restart every year over the 40 year lifetime of the plant) and could only be met by a reactor with a risk of releasing radiation to the public in the event of a maximum accident so small as to be functionally equivalent of zero.

What’s the actual result of this?

By making it so hard to build very very safe, very very clean reactors, the NRC is creating *real public health risks* from burning way more coal and gas than would be otherwise necessary.

“Public health risk associated with emissions of air toxins, particulates, and other dangerous pollutants allowed by EPA regulations of power plants under the Clean Air Act are several orders of magnitude greater than the standards that NRC proposes to enforce for advanced reactors.”

---

🎨 🎭 Liberty Studio 👩‍🎨 🎥

🎬 The Last of Us TV Show Trailer (coming January 2023)

I’m both excited and nervous about this one.

The game had such great characters and story that it’ll be hard to live up to the original and the expectations.

I don’t know what I think of the trailer. Part of me is disappointed that they don’t look and sound like in the game, but that obviously wasn’t possible.

I’m going to try to keep my expectations reasonable and hope for the best.

I’d really love for the TV show to be great, because the game is aging, and I know that at some point we’ll have a hard time getting past the graphics to get to the emotional core of the story. I’d love to have a more ‘timeless’ version.