American Dynamism or Manufacturing Stagnation?
Are we simply witnessing the pivoting of Silicon Valley spin doctors to hard engineering?
The star venture capitalists at Andreessen Horowitz seem to think that American dynamism is just around the corner. All America needs is a little optimism. Katherine Boyle, a partner at the VC firm, argued: “We believe technology is the font of this reversal, and the only immediate way to kickstart American renewal is through startups building for critical problems. But most importantly, dynamism requires optimism and a belief in growth and opportunity.”
Unfortunately, it is hard to be optimistic given the long list of headwinds facing any manufacturer in the U.S., which to Boyle’s credit she describes extensively. These challenges include: (1) government funding channels not focused on small firms, (2) supply chain issues in sourcing input materials and components, (3) lack of talent and skilled trades workers, (4) the only customer is the federal government. Some of these are spun as positives, but honestly they are all negatives.
Here’s how US manufacturing is going according to the surveys by the NY Federal Reserve and Dallas Federal Reserve (based on responses from their areas):
That’s 12 months of declining conditions and demand through May of 2023.
Add to this, the fact that the US is essentially at war with Russia (according to Senator Lindsey Graham and others). Normally a war against another superpower would spur manufacturing and demand, but not now. Instead we have to beg South Korea to sell us their artillery rounds, and then when our only factory manufacturing the explosives for artillery rounds self-destructed, we now have to beg Japan for TNT. I certainly hope our leaders aren’t serious about “freeing Taiwan”.
This environment would seem to suggest a huge demand for growth in manufacturing, and yet the graphs above show nothing of the sort. As they said in a past period of production and American dynamism: “you can’t beat a dead horse.” How to spur American dynamism when there’s no foundation? Our best electrical engineers, physicists, and chemical engineers have gone on to work in finance, software, or law. Our skilled rust belt workers have been told to learn to code and have been besieged by the opioid epidemic. This is the reality that silicon valley VCs have yet to deal with in their quest for industrial dynamism.
The silicon valley answer is that technology will be the solution. They say the US is a tech power house, just look at the S&P 500 compared to 1996. Well in 1996 the top companies were manufacturers using technology. Are we to believe that the Chinese are not using technology to remain competitive? Technology and AI cannot build the factory; it cannot get the permits; it cannot extract the raw materials.
In part technology is the problem. Technology has led us to believe that we can modularize, outsource, and assemble physical parts just like in software where modularization is the name of the game. Technology has pushed US manufacturing towards CNC machining of parts and away from casting and other mass production capabilities. CNC machining cannot scale like casting can, even with silicon valley money; nor can 3D printing. We have a US company trying to 3D print an entire rocket including the cylindrical body - sheer lunacy - but VC funded. In contrast, SpaceX has to teach their skilled welders how to handle their special steel formulation and quality standards by building dozens of prototypes. The road to scale in manufacturing is not build, point, and forget.
To return to manufacturing dynamism requires as a first step to embrace the manipulation of atoms at their fundamentals with, metallurgy, materials science, and chemistry. Then government and VCs must fund those who build and build cheaply. Here’s some criteria to eliminate spin doctors turned “technologist”:
If they only have a powerpoint, reject immediately. With technology today, most things can be prototyped with commercial off the shelf.
Reject pivots. This is not software, pivots in hard tech products are essentially a new investment. For example, Varda Space started as a satellite company, then was a rocket company talking about hypersonics, and now Varda Space intends to build factories in space. LMFAO. This doesn’t work.
Reject any funding plan that does not include a plan to iterate through at least 10 prototypes. James Dyson iterated through 5127 designs and tweaks in making his bagless vacuum. Failure in the world of atoms is two or three orders of magnitude higher than software, but the direction can sometimes be clearer. So funding plans can include performance benchmarks. This is why building cheap is important, iterations grow costs quickly.
Reject any product that is simply, “known product x made by robots”. If that is the company’s pitch and branding, they wont reject robots when doing so would be reasonable. See Relativity Space which 3D prints rocket fuselages. Similarly, avoid “connected” manufacturing type schemes like that described by AZ16: “companies such as Applied Intuition and Samsara are taking on the OEMs and Honeywells of the world by putting software and cheap sensors inside the industrial sector to improve safety and efficiency simultaneously.” This is a hack waiting to happen. Factories should not be connected to the internet any more than your refrigerator.