5 Comments
Sep 16, 2023Liked by Sarah Constantin

Neat! It's a nice insight that any technology that can take advantage of the massive R&D investment into chip manufacturing will have a big leg up.

Are you familiar with the 2020 paper proposing that the learning curve of a given technology can be predicted based on two factors, design complexity and need for customization? You can find it at https://www.sciencedirect.com/science/article/pii/S2542435120304402. I found it to be very helpful in understanding whether a technology is likely to succeed in the market, especially in the context of low-carbon technologies. (My own attempt at summarizing and explaining the ideas in the paper is here: https://climateer.substack.com/p/learning-curves.)

Expand full comment

What an eerie conincidence. I just finished reading [I, Exponential](https://www.notboring.co/p/i-exponential) on Not Boring, which shows the exact same three examples at the start: Moore's law, DNA sequencing costs, and solar power costs :)

Expand full comment

Great review - from the nitty gritty of computers to the big picture of technological progress. Thank you

Expand full comment

Simply brilliant! 💡

Expand full comment

Fantastic overview! So all the tech that improves exponentially does so because of semiconductor manufacturing, but why is it that semiconductor manufacturing has improved exponentially? How was it that for decades the crazy-sounding goal of doubling the number of transistors on a chip in 18 months for the same cost achievable?

Expand full comment