Creator of @Webchiver. Previously, founding engineer @HubSpot.
Oct 29 • 35 tweets • 13 min read
It’s time to rebut this completely, embarrassingly wrong and hostile thread. Link to article and long 🧵 below. Sorry for the delay, but quality writing takes longer than LLM-style writing that is superficially smart but uncomprehending and error-ridden.
- People admitting my points about manufacturing GDP real value-added even while calling me wrong
- A blow-by-blow rebuttal of Sichuan’s very bad article
- Technical details previously omitted for brevity
A few years ago I was surprised to come across economics statistics showing that U.S. manufacturing output had risen in the past decades, even as I noticed fewer and fewer things I owned were made in the U.S.A. That got me thinking ...
...how were those statistics even calculated? How do you add up "output" across a diverse array of complex goods? This past year I did a huge deep-dive on GDP manufacturing data methodology, digging through hefty handbooks and old academic papers...