Brownian motion

Article

Brownian motion is a recurring concept in the Astral Codex Ten archive, appearing 2 times across 2 issues between June 03, 2021 and November 18, 2021. The archive places it in contexts such as “Einstein’s annus mirabilis, when he published papers on Brownian motion”; “Einstein proposed that this Brownian motion resulted from the random movements of atoms”. It most often appears alongside Einstein, Scott, 23andme.

Metadata

  • Category: Concepts
  • Mention count: 2
  • Issue count: 2
  • First seen: June 03, 2021
  • Last seen: November 18, 2021

Appears In

Source Context

Recovered passages from the original issue text. When the raw archive preserved outbound links inside the source passage, they are listed directly under the quote.

June 03, 2021 · Original source
When a field is established, you can get a lot done by making relatively obvious inferences and running cheap tests. (I’m not sure what share of the patents in the above graph were physics-related but it seems notable that 1905 was Einstein’s annus mirabilis, when he published papers on the photoelectric effect, Brownian motion, and special relativity). After that, you reach the questions you can only resolve by smashing particles into each other in a 10-kilometer long tube, or by buying millions of dollars’ worth of GPUs so you can check how much better neural networks do if you just let them have a billion parameters.
November 18, 2021 · Original source
The trouble is that this isn't how science worked, back when it worked. One person would notice something funny--not necessarily a theoretical problem; often a pure disinterested observation of just the type Popper claims is impossible, such as when Robert Brown published a paper basically saying "Microscopic particles in my tea keep jostling around as if they were alive". Another person might propose a hypothesis, as when Einstein proposed that this Brownian motion resulted from the random movements of atoms. Then a third person might propose a test, and a fourth might conduct the proposed experiment and report the results. This entire process sometimes took a century or more.