7 Comments

Thank you for the great blog series! Regarding footnote 2, "Today, I’m going to skip over the philosophy of why we’ve decided to trust these algorithmic reconstructions of reality.", do you have any pointers for reading material on this/what terms I should search for?

Expand full comment

A classic is Daston and Galison's "Image of Objectivity."

https://www.jstor.org/stable/2928741

Expand full comment

By constant c in the last paragraph you're referring to the noise-image tradeoff term, which is different from the c_i coefficients in the equation right?

Expand full comment

That's right. I just changed c to h to avoid this confusion.

Expand full comment

"For example, the measured output in a CT scan is the amount of X-rays that made it through the body. This decreases linearly in the amount of absorbing material between the X-ray emitter and detector."

Surely only over some range? My lead blanket stops (nearly) all the X-rays at the dentist, right? I shouldn't ask for a second lead blanket to stop twice as many X-rays?

Expand full comment

yes, only over some range.

but maybe you should ask for that 2nd blanket at the dentist? Dentists are too into X-rays.

Expand full comment

I proudly work on inverse problems (and I love atomic norms)!

Expand full comment