63 Comments
Jun 11Liked by Noah Smith

"Which is why Democrats need to stand up and lead."

Noah's conclusion says it all. The Trump GOP is, sadly, not up to the task. Despite their political rhetoric to the contrary. Their latest natsec betrayal was Speaker Johnson's six month obstruction of a $90bn investment in American munition manufacturing capacity. That's SIX months. At Trump's urging. With friends like that, who needs enemies?

Expand full comment

And the Democrats wouldn't have passed a bill to invest $90bn in munitions manufacturing.

Expand full comment

Yeah, the folks who get their news from FOX and the mini-FOX's were told it was all some huge give-away to Ukraine. When it's actually a sizable investment in our own military, after which we give Ukraine our cast-offs to fight Communist China's newest little buddy.

Expand full comment

ABC, CBS, NBC, CNN, et al, called it the Ukraine Aid Bill, though.

Expand full comment

Sure. It's a help-Ukraine-but-help-the-USA-more bill. China's...errr, Putin's invasion of Ukraine was simply the catalyst for a move to modernize out own stocks, and boost munition manufacturing capacity.

Expand full comment

True, but... they really shouldn't have lied about it.

Expand full comment
Jun 11·edited Jun 11

Politicians create all sorts of misleading names for proposed legislation...like the Healthy Forests Act (to clear-cut forests), and do on. When it comes to creating names for legislation, Congress is like a retarded version of Madison Avenue...they're pretty lame. On both sides of the aisle.

But the GOP, at Trump's urging, blocked that bill precisely because it would help Ukraine, knowing full well it would help our own military MORE. The very definition of traitor, no?

Expand full comment

Let me make sure I understand: Democrats are now claiming that they are the ones who have always supported military expansion. Talk about revisionist history!

Expand full comment

You've got it wrong. The GOP has historically been the militarist party. But since Bush II they've completely dropped the ball on modernizing the military to face China, and balancing the budget (by modernizing the IRS, ending GOP cuts to taxing the rich and Big Business).

But the Trump/Johnson spiking of the $90bn "Ukraine" bill says it all. A $90bn expansion of our munitions manufacturing capacity--that the GOP would have permanently spiked if they could have.

Expand full comment

Well, THAT doesn't help. Bush II was criticized (by people like you?) for invading Iraq and further establishing a military presence in the Middle East. Arguably, he spent more and sent more troops to the Middle East than any other president ever.

Expand full comment
Jun 12·edited Jun 12

The Bush II wars added up to $7trn to the national debt, according to some estimates. And for what? Shia Iraq is now more closely aligned to Iran than us. The useless and unsustainable forever war in Afghanistan had to be (finally) ended. Up to 2 million Iraqi refugees from our war flooded Europe, weakening NATO and destabilizing the region.

Could that $7trn--or even $4 to $5trn--have been better spent beefing up the Navy and AF in the China theatre, instead? Damn right it could've. Flawed GOP leadership expended massive amounts of US treasure in these mostly useless military endeavors in the Middle East, when even the Pentagon was urging the money to be spent on containing China.

Expand full comment

Wrong

Expand full comment

I don't get this. I recall Trump being the only candidate in 2016 who was talking about China, and everyone else called him racist for it. Now he's the candidate that's weak on China? How did that happen?

Expand full comment

He was called racist because of later comments like "Kung-Flu" and "China Virus." But DJT did ruffle GOP natsec feathers in 2016 by rubbing in the fact they'd wasted $6 TRILLION dollars on their failed Bush II wars in Iraq and the 'Stan.

Expand full comment
Jun 11·edited Jun 11Liked by Noah Smith

I built radio control airplanes 30 years ago. 6 ft wingspan. 20lbs. And I worked with the auto industry for 20 years. Been in every plant that makes everything you can think. Steel. Dies. Stamping. Copper smelt, wire drawing, wire coating. Motors. Actuators. Glass to optics. Every plastic. Aluminum. Forging, progessive die stamping. Die casting. Blow molding, extrusion.

So like a car system, a drone needs a few basic things in high volume.

Mechanics and structure - light weight injection molded plastics. Fuselage .empennage

Electrical system:

Brain - CPU - make or ask Samsung. Buy a million Androids. Use them

Sensors. CMOS - cellphones

Optical lenses- cellphones again. Injection

molded or high volume glass molding

Actuators- motors- high speed, small servos. Moog & auto electronics

Gyros- well ..develop that here

Flight. Sense. Compute. Act-actuate.

Basics -

A. This is absolutely not forever something low volume, high cost defense industry can do..

B. This is not good for VC startups. They dont make many things. Too much learning curve

C. Between teams from Detroit, a bit of current DoD drone experts, Mexico or US volume assembly, and Samsung - you have your China beating DDT - Drone Dream Team.

14 months - 100,000 6ft drones a week. 1 million 1 ft drones a week. 100 20ft strategic drones a week

Book it.

Expand full comment
author

Love it.

Expand full comment

🐾🐾🐾 🐰

Expand full comment

You left out the radio components which are needed for command and control, as well as reconnaissance.Although for military purposes a drone needs to handle radio frequency jamming and GPS spoofing, you would need spread spectrum or other jam resistant radios and an alternative for GPS (with some kind of battlefield local positioning from an EWACS like mother drone to combat spoofing). For targeting and signal analysis a stronger ML based processor is probably necessary, as well as IR cameras and maybe LIDAR.

To act autonomously as swarms (as demonstrated by the nighttime drone displays at festivals ) the drones need to have secure intra vehicle communications and AI for controlling behaviors. Software is a big issue, Anduril is working on making autonomous swarming drones for the US military now, but there isn’t much drone software experience here either. Also missing on your list is batteries which also come now mainly from China.

Expand full comment

This is why the StarLink system exists.

Expand full comment

Excellent points. I mentioned Communications somewhere I thought. But yes, secure radio needed. Although I think preloaded AI/algorithm might allow the swarming drones you mention.

You ready to start ! :)

Expand full comment

Love this. People gripe about General Motors, but they were up making ventilators during COVID at a speed that was very impressive.

Expand full comment

What’s a rough order pf magnitude price tag for this?

Expand full comment
Jun 11·edited Jun 11

I would estimate a rough order of magnitude to achieve the 14 month goal as:

Design Engineering: Including Electrical/Electronic; Controls, Optics, Flight/aero, Comms; Payload (military does it) - $25 million. Built around the automotive APQP (Advanced Product and Process Quality Planning), with adds/inputs from Samsung Electronics and motors. Engineering development lines. FTE, management, investment for Engineering.

Program Management, Systems Engineering, Supply chain. $5M

Manufacturing Engineering: Tool design, Process Design, DFMA. Pilot lines. And scale in steps to full production. $25 million.

Production Tooling and Line CAPEX:

Depends on Samsung COTS, or new efficient integrated electronics. Integrate CPU, Sensor collection, flight controls, gyro/GPS/Rad resistant electronics.

$100 million to $500 million.

Expand full comment
Jun 11Liked by Noah Smith

"One idea here would be to reward strategically important companies in allied countries with access to the U.S. market in exchange for harmonizing their China trade policies with America’s." When so many of the Fortune 500 companies are already reliant on the Chinese market, I doubt corporate leaders of companies in strategic sectors would fall in line and might only offer lip service, no?

https://www.shanghai.gov.cn/nw48081/20230804/2d49914657a74258a4a010346c5f4b28.html

Expand full comment
Jun 11Liked by Noah Smith

“That drone was piloted remotely by a human, but in the not-so-distant future, drones will be autonomous AI-powered killing machines.”

Not to get to effective altruist-y, but if turns out to be true, wouldn’t this arguably be a greater threat than China? Even ignoring superalignment issues, imagine if some terrorist group or rogue state gets their hands on this type of technology.

Expand full comment
author

Drones are really capital-intensive -- you have to produce a ton of them, keep producing a ton of them as individual ones are lost, and you also have to support them with a huge back-end infrastructure. Terrorist groups don't really have the capacity for any of those things -- even the Houthis had to get their drones from Iran.

What I do worry about is terrorists using single autonomous drones to assassinate people using facial recognition.

Expand full comment
Jun 11Liked by Noah Smith

Noah's point is that those swarms of autonomous military drones will be PLA-controlled drones. NOT American ones. China is already as advanced in AI computing as the US. And we have roughly zero drone manufacturing capability; while China leads the world.

Expand full comment

Are you not at all familiar with ISO-9000? I also suspect you have never worked at a manufacturing company before either. Tracking where the chips in a car are from should be a snap at any reputable company. I consulted for Nucor Steel to get them ISO-9000 registration because their customers, and their customers, …, insisted on traceability of the chemistry of their steel. Chips are much easier to track.

Expand full comment

But nobody seems to collect these data from many companies into a single useable database.

Expand full comment

It’s not just cars, there are hundreds of thousands of imported products with chips right now on Amazon, for some of them nobody knows if they even have a chip in it, let alone where it comes from.

Expand full comment

I keep reading that China’s exports are unstoppable because the Chinese government subsidizes them so much. But then I also keep reading about how China’s economy has problems with low internal consumption and that exports are the only real source of growth for the country. These two ideas seem contradictory. How can the subsidies keep flowing into exports if exports are the only real source of revenue? How long can China maintain this pace of export led growth?

Expand full comment

It's a throughput based model. They give bottomless loans to most anyone who can show they are employing people. Someday the bottom will fall out (maybe already has). But for now they keep cooking their own books harder and harder every year. Ultimately they have mortgaged their future so hard that I think they are 3x underwater as a nation (if one could liquidate a country without crashing the prices of everything they would still own two extra years of GDP to the debtors after). So that can go on until they run out of For-ex and all of the countries that import raw materials or food to them stop sending shipments. Then they all starve, it's going to be ugly and I think we need to prepare to feed them not kill them. Cause they're gonna need the food.

Expand full comment

Thanks. If this is the case, then wouldn’t the best grand strategy for the US be to keep buying their cheap subsidized stuff in order to accelerate their economic collapse? Maybe the people thinking about this great power competition don’t actually believe this theory that China is economically underwater as a nation - Noah certainly seems to think they are immune to economic problems.

Expand full comment

Yes, there's an argument for that. If China subsidizes drone manufacture then OK, buy a ton of super cheap drones from them and stockpile them. You end up ahead, they're basically paying for weapons that can be used against them. On something as simple as a drone the software is not really an issue to worry about w.r.t. backdoors (it can be easily replaced).

The risk is that you can't build up your own drone manufacturing plants that way. But, is that really a big risk? Chinese are certainly adept at quickly cloning foreigner owned factories and tech, why would the west be any slower?

Expand full comment

IT kinda a matter of philosophy. If you are willing to shoot anyone who goes to the bank to riot when the housing collapse comes then you may be able to avoid the currency collapse. But it's probably a matter of volume and degree.

Like how many cops want to arrest someone who stopped paying on an underwater mortgage of a second home when their grandmother just did the same thing?

What is "money" anyway? It might be a decent proxy for trust in a system. If so then printing it (endless loans) just dilutes it and can't create productive capacity.

To do the subject of investment led growth justice takes multiple gigabytes of spreadsheets. I think that China is broke like the EU is broke (no-one wanted to admit the Greeks were bankrupt because doing so might light off the financial crisis in Italy so they just kept giving the Greeks more money to kick the can down the road and let the next parliament deal with it) .

But what I think is unimportant what matters is how far can the populations standard of living shrink before they say F-it I'm done paying taxes? It happened to the soviets in 1989 will it happen to the CCP in 2025?

Expand full comment

“At that point, it’s an open question of whether human soldiers will even be able to survive on the front lines of war.”

I take it you have never shot a shotgun before, either. Anyone who has ever shot skeet, trap, or just hunted could knock down that UAV pretty easily. Plus, counter measures beget counter-counter measures beget…. Someone will always figure out a way to defend against a new weapon.

As far as the video being stomachable, I think it should be put to Benny Hill music and broadcast throughout Russia. That should help kill their morale.

Expand full comment

Most of the people in Russian military come under pressure (economic or... well, otherwise), their morale is already very low, but they are generally afraid of their own officers even more.

Expand full comment

Forget hand guns . A drone with a bomb in every home . It’s part of the constitution . We’ll fight till the last pronoun . Better dead than red . Where do I invest in the military industrial complex ?

Expand full comment

The problem is Biden hasn't really decided if this is about narrow defense issues or about more general economic competition.

Protecting our key defense industries makes perfect sense but generally trying to limit China's ability to sell us cheap junk and stop them from growing richer doesn't.

Expand full comment

This a damn good read. Three points:

1. In the age of almost unlimited data there is no excuse for flying blind on imported tech country of origin components.

2. Tariffs are good weapons to protect nascent and strategic industries as well as to PUNISH unfair trade such as technology thievery.

3. US military is woefully too small and diluted given our history, magnitude of potential military foes, and the extremely high likelihood of a two theater war. The tooth to tail ratio right now is the worst it has been in U.S. history.

The military has not routinely met recruiting goals in a decade.

JLM

www themusingsofthebigredcar.com

Expand full comment

I agree we’re living through an amazing moment of data abundance, yet, collecting data on the chips that power imports from other countries doesn’t seem *that* easy to me. It’s going to take high levels of coordination with a lot of different players 🤷🏻‍♂️

Expand full comment

Interestingly there is no mention of a significant carrot and stick the United States has with LNG which the Biden administration has illogically sided with the climate extremists to restrict. Natural gas will be a large component of heating and electrical generation and industrial use for decades. Some important economic allies, like Japan, South Korea, Germany, France, Spain, Taiwan, UK, and India (potential ally) are already big importers of LNG, while a place on the fence, like Vietnam, is building LNG import capability. China is huge importer of LNG. Speak softly and carry a big stick. For the next couple decades that describes LNG. Large parts of the world need natural gas for some time and the US has the supply. Who gets LNG, at least initially, can be exactly controlled by the US, a combined carrot and stick.

I grew up, as much of my extended family has on the SF Peninsula, and benefitted from what was one of the great industrial policies we have seen: The Cold War. This was industrial policy by carrot, the military wanted and funded increased research to beat anticipated rivals with better technology, but spent far more on buying the military products and support it contracted for. The NASA Space Program also had a big impact on the Peninsula as a large new consumer of cutting edge electronics developed and built on the S. Peninsula.

More drone capability, at different levels of sensor sophistication, range, and armaments, appears at present to be a military requirement. What has worked in the past here is having the US military buy a steady supply of something where in this case there is a requirement that it is controlled by a single chip, designed and produced in the US at a contract FAB and an similarly produced sensor management chip. Could even do similar things with motors since likely motors will have embedded digital controls and require them to have only US designed and produced digital components. DC electric motors are readily amenable to robotic building.

Suppose the US military agrees to buy 1 million $5k drones/year and 250K $20 drones/year, subject to the domestic requirements above, for the foreseeable future. That’s a small fraction of the military budget at ca. $9B/year vs. the military budget, currently around $900B, but certainly getting larger (e.g. the rightward shift of the electorate in nearly every, but not all, democratic countries post covid). So to make a difference you probably need to be at least an order of magnitude larger. The problem with including much higher priced drones with the same restrictions is what is the consumer demand for products resulting from this development. There is some industry demand at higher price and performance levels, if you say had drones instead of helicopters for lots of observation. Companies could monitor things like electrical transmission lines, gas pipelines, methane emmisions, …, with higher end, longer distance, more sensors, drones. Commercial drones could revolutionize backcountry skiing, allowing the accurate dropping of charges even by individual groups, but probably more important for resort and highway avalanche control.

Kalecki, using a Phillips Curve argument, argued that a certain type of government spending on armaments allowed for full employment in budding nominal autocratic right wing governments which was accepted by the industrial leaders who benefitted from government armaments contracts. By the numbers we were in the forefront of military spending after WWII much of the time. The Hanford Site, where Plutonium was created in high flux reactors for the FatBoy bomb and most of the thermonuclear devices we still employ, is one of the greatest human engineering feats in modern history.

Expand full comment

FWIW given our last interaction on Twitter on the topic, your point on trade is the one I want to hammer in. Right now we seem to just be moving in a general protectionist direction across both parties, when we need to be economically integrating MORE with other OECD countries.

Expand full comment

Viewing all problems through a political lens is limiting. Most companies dont have a clear picture of their supply chains 3 or 4 levels down. The mining and processing of critical minerals and transportation costs are often times overlooked. Government is neither good nor bad. Most often, it is a marginal player on the outside of a complex system with a limited ability to influence the system. Co-development and co-production are serious efforts that can take decades to play out and will pit policy goals against each other. More often than not, decisions end up being value trade offs, right vs right, than the right vs wrong that our political lens nudges us toward absolutes. First step, stop speaking about aspirational goals as if they've already come to pass and start working soon practically achievable goals that involve the markets, workforce and governments

Expand full comment

Interesting article... I think a little off-base....

1) On Legacy Semiconductors, this is a temporary issue to be managed. How did we get here ? We have non-consumer markets which have system products which last decades (DoD being prime example). They have older designs which consume mostly standard parts which were non-Chinese sourced for many years. The US/Europe got out of these markets. Why? Leading edge chips are order-of-magnitude cheaper, faster, lower-power, etc... all at the same time. Noone... (except China ..driven by the intense need to break into the business) wants to invest in legacy fabs. This problem gets naturally solved in two ways: 1) Small number of fabs in friendly countries to maintain lagging supply 2) system redesigns to leading edge nodes. Both of these are happening. Note.. it is going to be interesting what the Chinese are going to do with all the legacy fab supply. I suspect another example of massive misallocation of capital.

2) Drones.... the primary issue is cost ... which either gets solved by another low-cost location or automation.

3) "It’s important to remember how much bigger China is than the U.S. It has four times the population and twice the manufacturing capacity by value-added. Any protracted conflict between the two countries would see the U.S. badly outmatched, despite its modest remaining advantages in some areas of technology." .... wow... this really seems off base. The population doesn't matter.. India/Africa are bigger. On Manufacuring value-added, maybe.. but their supply chain dependency is way bigger than the US. Also, they are flooding world markets in some very specific industries with over investment (EV, legacy Semi, Steel, etc).

Overall, the root of all the challenges from China come from their ability to build at a much lower cost point. This is largely based on labor costs (real...gdp/capita), currency manipulation (artifically pegged to dollar), and economies-of-scale in certain verticals (construction, electronics assembly, etc) [real]. If I was the US, the focus on somewhat controlling the leading-edge is good, but the other focus should be obtaining a diversity of supply for lower cost manufacturing. The obvious answer is Mexico/Central America, but perhaps also India/SE Asia.

Expand full comment

The fourth hole is the biggest of all: the fiscal deficit that makes all US exports less competitive and all imports more competitive. Besides defects suck resource from investment in drones or chips or whatever.

Expand full comment

A deficit greatly magnified by the Fed's rate hikes...that are supposed to lower inflation by slowing the economy, and/or causing an outright recession/depression; but which also raise inflation by increasing rents (due to nixing new housing starts), and expanding deficit spending.

Macro-economists seem to get so many things wrong. Who is to say whether a Fed-led global round of central bank rate cutting would actually create a net increase in inflation?

Expand full comment

Why not ban all trade with China, as with Cuba? Why not a general rule that the US does not trade with non-democracies?

Expand full comment

You'd first piss of the ME and bring down decades old f American efforts there. You'd have calls from nearly every Fortune 500 company in the world Lobbying against it. You'd also have calls from Brussels wondering if Biden IS out of his mind. Same thing with Japan and Korea. China is not the Soviet Union. Being an economic and trade powerhouse means that Beijing is far from defenceless. Also, that can escalate dangerously. At that point, China's trade ministry will pull out lots of plugs and then the US will be forced to retaliate. Soon someone in America will get the stupid idea of removing the world's largest trading nation from SWIFT in peacetime. It'll also change the narrative from, we're trying to contain Chinese defense advancement, to we're actively trying to make Chinese people poorer

Expand full comment

Because Intel, Apple, and a number of other Fortune 500 companies would go absolutely ape-shit to prevent that. Wall Street created the ravening beast that is now China; but don't expect them to pay for their mistake by actually paying for it.

Expand full comment