Random Thoughts – Page 2

AI’s $600B Question

Obviously it’s a question that needs to be addressed: where are the returns on Nvidia GPUs?

Here are some of my thoughts on what’s missing from the article – not to say I have an answer, but to look at the article/question from other perspectives

1/ what are the risks of LOSING revenues if falling behind in AI?

Case in point – likely that Google can’t keep its grip on the AI equivalent of iOS search engine bar (or whatever the gateway to AI functions and monetization) in the next decade.

2/ what are the risks of losing top talents’ interest if not doing AI? And the culture of being at the frontier?

People wants to be part of the next gen thing. To be precise, top talents want to be the center of the next gen thing. If they see peers doing AI, they will not forgive themselves not doing AI.

If the company is seen as not the frontier, that’s a big risk in the next decade of not getting the top talents naturally. I bet now Google needs to do EXTRA to get talents it wants vs in the early days they were drawn to Google.

3/ what are the “infrastructure”?

In the article, author mentioned railroads. So what exactly is the “railroad” now?

The GPUs?

Or the data centers?

Or the foundation models?

Where does the overinvestment risk lie? (I bet it’s the third; also DC to some extend)

4/ If there any “moral hazard” here?

By having investments in AI, VCs should be inclined to discourage similar investments from other people.

Few entrants would enhance their return – whether it’s foundation models or GPU resources (easier to get GPUs / falling prices).

It’s tricky.

Like inflation – you need to tell the public to spend less to lower inflation; if you tell the public that you expect inflation to be down too early, it would be harder to come down.

If you are telling people AI is a good investment, then it may make returns lower.

Reasons why can’t go full BEV in the near future

Few years ago, many car companies were committing to go full EV. Now it’s clear that won’t be the case.

Take Benz for example, in 2021 it stated it would go full EV for new models from 2025 onward and all EV by 2030 where market conditions allow. Now in 2024 it back pedaled, saying by 2030 will only do 50% EV.

Why?

– The most common reason is weak consumer demand due to weak charging infrastructure plus ICE outperforms in many use cases.

– EVs are not cheap enough.

What’s more?

Many countries don’t want to rely on China’s supply chain.

And?

There are deeper implications/concerns.

Auto industry in EU is built on ICE cars. The traditional car industry provides jobs, income, taxes, etc.

The stability of EU relies on traditional vehicles!

Btw, this also affects Japan with the same logic.

Auto industry is a smaller part of US economy directly, and US has Tesla, but US will likely be negatively affected if EU and Japan is unstable or poorer.

Therefore, either EU/JP needs to maintain competitiveness in the EV era, or they need to pivot to other industries (very hard; nothing in sight), or US will need to strengthen its own economy and rely less on EU/JP.

Why overcapacity is prevalent in China

Thinking about this question, here are five potential answers.

1/ For the older generations, shortages are deeply rooted in their memory, so they dislike the idea that anything might be in short supply.

2/ The required return is too low. Many experienced poverty in their youth, therefore lots of effort for little money is still attractive.

3/ The local governments want to create local jobs and boast about industrial upgrades, with banks being the main enablers. When every local government invest in a similar direction at the same time, it’s hard to control and estimate the entire capacity.

4/ For top party members, abundance of all sorts of goods is a sign of victory for socialism and communism.

5/ China is preparing for war. Many capacity will shut down if at war.

Boeing and US manufacturing

It’s hard to be top notch in every industry.

The best talents in a country will bring strength to one industry which is tech for the US. If tech gets the smartest minds, of course other industries may not be able to compete at their max potential.

Boeing pleads guilty for 737 Max crashes.

Boeing benefited from its market position and gov contracts, while GM needs to compete in the more fierce commercial world. However the drivers and outcome feel a bit similar to me.

It could be about global competition but that’s limited (Airbus and SpaceX), and Covid is just one-off. Blame China? Sure China is not buying many planes like before, but it’s very early in actually competing in airplane manufacturing. I assume Airbus is not breaking many rules to compete. SpaceX might have some technology breakthrough, but Boeing also has r&d and should have unparalleled industry expertise.

The harder question is what’s Boeing’s (and broadly US manufacturing) competitive advantage over the longer term? Why TSMC didn’t do well in the US?

From a previous post, I touched on this – how to recreate the manufacturing industry in US is a question about how to make this sector attractive compared with other industries for young people and for capital.

Fiscal spending and subsidy is not a permanent solution. Tech didn’t rely much on subsidy to grow.

Maybe there is a negative side effect of being very strong in the technology sector?

Maybe US has evolved over the years so that the nature of this industry didn’t deserve a lot of attention? Maybe Tesla robot is an answer to unleash the manufacturing potential.

Small countries need a union

It’s tough for small countries to negotiate with big countries, like any single individual is hard to fight a big company.

To protect individual employees, union is needed.

It’s the same for small countries.

It’s different from UN, which INCLUDE big counties. It’s not like BRICS etc., as China is already a super power.

It should be small countries only. And they shall work together and organize a team to fight for their rights in negotiations.

Sure this may cost some money, but the benefit shall outweigh.

They shall need to resist the attempt to make separate deals with big countries – this will undermine the coalition and bargain power collectively.

Indirect casualties

Prolonged worsening US-China relationship has more indirect casualties it appears.

Recently, a women who defended Japanese nationals from being attacked in Suzhou passed away.

A 30-year girl in Shanghai suicided when facing a declining housing price (~3-4mn rmb of down payment loss) and cut in pay.

We might not feel US-China relationship in our daily life, but it’s around us.

It’s affecting every corner of the world.

Bad is good. Need some losers.

Nike stock was a loser last week. Nike is under pressure with fiscal year 2025 to be down mid-single digit in revenue.

However, this is exactly what the US needs right now – weaker consumer, cooler economy.

Lower demand tempers inflation and it’s easier for Fed to see the data it wants.

When Fed gets what it needs and cuts the interest rate, others thing could look for bottoms or better terms – whether it’s commercial real estate or refinancing some low cost CBs.

If you want to get there faster, you better hope for the opposite.

When more people see only one cut or no cut, overall economy will cool a bit. Companies and consumers would tighten their belts, which will be helpful to the Fed.

When the market is pricing in “higher for longer”, multiples come down & stocks drop. Here, the reverse wealth effect will drive down inflation.

Talk bad stuff, so you can get good result in the end.

Liquor (Baijiu) consumption in China

I came across this number today: 6.29 billion liter of Baijiu (a traditional Chinese liquor) was produced in 2023.

I am wondering how to make sense of this.

1 liter = ~0.9kg or 18 liang (两, a traditional Chinese measurement)

6.29 billion liter = 4.5 liter per person, or 81 liang

So if all produced were consumed, it’s 81 liang per person per year, or 1.5 liang per week.

If you assume children (250mn below 15) and many women don’t drink Baijiu that much, it could 3 liang per adult male per week.

It sounds like a pretty high number. One typically won’t drink 3 liang if eat at home. It’s usually for dinning out with business or other occasions.

Considering how occasions like this would decrease over time – younger generation might prefer a different life style, or businesses could be done without being drunk at dinners, it’s hard not to question about how much Baijiu China needs to consume each year.

I won’t be surprised if the volume can be down 1/3 or 1/2 from here in decades, assuming no export or store of value etc.

About printing money

In the US, when the US gov borrows money, the treasury department will issue different debt securities with different maturity and interest rate.

When demand is low, and interest rate may shoot up. Meanwhile, the Fed may step in to buy treasuries in the open market, by crediting (increasing) reserve accounts banks hold at the Fed. Also Fed needs to pay interest for these reserve accounts, it doesn’t really matter that much – the loss Fed made can be earned in the future. In normal times, the Fed’s profit will go to the treasury department.

So everything looks like magic and money is just created from the air.

The only process that needs the public to participate is the US treasury auction and the Fed can’t buy directly.

In China, property looks just the same.

When a city gov borrows money, it will sell the use right of a land. If that’s for residential, then they will be held by homebuyers eventually. When demand is low and deficit is high, home price shall drop.

So developers are like the banks participate in US treasury auction. But who is the Fed in this case?

It’s interest that China has thought about the idea of buying back unsold homes.

In some cases, homes will be auctioned by banks if the homebuyers default. Anyways, there is excess supply on the market and demand is not strong.

The problem is this is not central gov’s deficits. Local gov doesn’t have a “Fed”. Maybe some local SOEs can act like Fed in this case to buy properties back?

But these local SOEs can’t just buy by creating “reserve accounts” at local banks.

Sure they must have good relations. So basically local banks need to lend to local SOEs and not to worry about these “reserve accounts”. Put it another way, banks need to swap the homes (treasuries) with local SOEs promises (reserve accounts).

You see, this has become very complicated and is not as smooth as printing money in the US.

What if 四大名著 are four major game types?

Just a random thought..

西游记: this is easy – RPG game with many bosses battles. The essence of the book is a group of 4 ppl going the West with many monsters on the road.

三国演义: this is easy – SLG strategy game (turn-based), or RTS game. The book is about a period of Chinese ancient history when three kingdoms fight each other.

水浒传: this can be gacha game like Genshin Impact. The book has 108 characters, each has unique backgrounds and skills.

红楼梦: this one is hard. What makes this book unique is its detailed depiction of lifestyle during Qing Dynasty. Maybe it’s metaverse game, or story-based romance game.

These 4 famous books each has its own appeal to a wide range of audience, which can reflect certain underlying demands. And these books’ popularity is proven over time.

Game developers need to constantly think about what “needs” they are catering to. Any thing that is popular for a long time must have served some pain points well.