Your correspondent is highly sympathetic to the proposition that lithium battery electric vehicles are, to use a highly oversimplistic and not-at-all-controversial term, a scam. While battery efficiency has been improving, there is good physics to suggest that there is a hard limit and they will never get to the point where they can afford as much range and convenience as gas cars. More to the point, given their supposed environmental benefits, there is a lot of evidence to suggest that the manufacturing process of the batteries makes them on balance more environmentally damaging than gas cars. 

And yet, we may find in some possible semi-distant future that Tesla Motors contributed enormously to humanity, even as its battery-operated vehicles fail. An obvious way is in the area of autonomous vehicle. 

But an other, less-understood way is in the area of manufacturing. Elon Musk decided he wanted to build hundreds of thousands of electric vehicles. Every car company said that was impossible. The way he solved this problem was the gigafactories: a giant battery factory, with a giant car factory right next to it. Because those factories were built from the ground up, they could have no legacy systems, could be built from the start to be software-operated, fully modular, and as robotized as possible with state-of-the-art technology. With all of that, Musk was able to manufacture his hundreds of thousands of cars, do it in a cost-competitive way--and create hundreds of thousands of jobs in the process. In the process of making electric cars, Musk may have invented a new way to do manufacturing competitively in a first-world country.

Yesterday at the Reboot Conference, we watched how Trae Stephens, chairman of Anduril, the biggest and most visible of the new crop of defense technology startups, described how Anduril poached the executive who built Tesla's gigafactories with the goal of building something they call the Arsenal of Democracy: a gigafactory for building advanced defense equipment—and not necessarily just for Anduril. We have seen in the war in Ukraine that 21st century warfare is extremely material-intensive and that it very much depends on a manufacturing capacity to build very large quantities of ammunitions, artillery, drones, and so on. A manufacturing capacity that the US abandoned decades ago. A manufacturing capacity that the US may yet regain thanks to Anduril. And it couldn't have happened without Tesla and its electric cars.

Who could have predicted this? 

No one. Not even Musk himself. 

Here is the point we took from this observation, and it's a basic hayekian point on the limits of knowledge. You should let the great men of history work. Let him cook, as the kids say. Even if you think his idea is stupid. Even if his idea really is objectively stupid—it may still have downstream second and third-order effects that nobody anticipates. 

Western civilization used to be notorious for its production of such men, men with crazy schemes and the capacity to put them together. Some call them the great men of history. Western civilization has become much more risk-averse and committee-driven. There must be some ways for policy to help let the great men of history cook.

Policy News You Need To Know

#PermittingReform – Yesterday () we gave you an overview of the House version of permitting reform. In the meantime IFP Infrastructure Fellow Aidan Mackenzie has come out with a great thread detailing what's in the bill.

#Trade – The Biden Administration is going to block Nippon Steel's acquisition of US Steel, WaPo's Jeff Stein reports. Probably inevitable in an election year where the candidate is fighting over manufacturing states. 

#Immigration – CIS has a good overview of Sen. Kamala Harris's record on immigration. Spoiler alert: it's not good. 

#Immigration – Speaking of, Rep. Chip Roy's office has put out a new report on the border crisis in the past few years. It's excellent.

#AmericanManufacturing - “China, Korea, and Japan together build over 90% of the world's shipping tonnage. The US builds just 0.2%,” reports IFP Senior Infrastructure Fellow Brian Potter. Why can't the US build ships?, he asks in his newsletter, Construction Physics, which we'll be subscribing to. A vital question. The article attributes this persistent lack of competitiveness to several factors, including high labor and material costs, insufficient motivation to improve efficiency and adopt new technologies, and protectionist policies that sheltered the industry rather than driving it to become more competitive.

#AmericanManufacturing – This is the best argument against protectionism, and one that protectionism advocates must address better: Intel, once the world's leading semiconductor company, has missed several key trends in the industry for a decade or two now, from mobile to advanced GPUs for AI, and has been correspondingly declining. After posting a particularly disastrous earnings report, the company is looking for what is in effect a bailout under the CHIPS Act, reports Bloomberg's Mackenzie Hawkins. Intel is requesting “up to an $8.5b grant & $11b loan,” and it wants it “ASAP.” Obviously, the CHIPS Act was intended to get the leading companies to build advanced semiconductor manufacturing facilities on US soil to support national security objectives in the competition with China, not to bail out underperformers. But in many cases this is what happens in practice with these types of policies. 

#DEI – The University of South Carolina required all students to affirm the value of "diversity and inclusion" as part of a mandatory training this summer, reports the excellent journalist Aaron Sibarium. “Then, when I reached out for comment, USC claimed the training was 'optional' despite telling students it was 'required,'” he writes. This un-American garbage is taxpayer-funded.

#AI – Quote of the Day from Palantir CEO David Karp, on large language models: "They are incredibly capable probabilistic engines that nonetheless have little innate sense of north or south and struggle on their own to solve complex analytical problems. Such models are necessary but not sufficient. Something more is required. They are wild animals, whose power and capabilities must be tamed and harnessed. And we are now seeing what is possible once they are."