Being made in a factory is important but I think to get a steep learning curve you need both ongoing manufacturing and design improvements.  For example, today's solar panels are not just less expensive.  They produce more power per unit area and last longer than ones made just a few years ago.

- High enough volume to justify tooling and automation

- Profitable enough to pay for R&D

- Base technology implementation is immature enough to provide room for improvement (from the start we knew the theoretical efficiency limit of silicon PV was nearly 30%)

- Design improvements can be fit into existing production flow so you don't need to recapitalize and relearn your manufacturing lessons.

- Regulatory and safety do not prevent ongoing changes and improvements to the design and manufacturing (will regulators let modular nuclear designers continuously improve their designs?)

Expand full comment

Great episode! There was a really good paper in 2013 that looked at this issue from a similar perspective, although using a somewhat different methodological approach. I think the two really complement each other so well worth checking out Dahlgren et al (2013) Small Modular Infrastructure (https://www.tandfonline.com/doi/abs/10.1080/0013791X.2013.825038)

Expand full comment

Sounds like the learning curve predictions for what will get cheapest fastest are awfully close to: "how much can be done in a factory?"

Expand full comment