In my recent episode, Ep 41 Knowing about Knowing, with Brendan Howard, we continued our discussion about knowing. Our conversation was based on a survey about epistemology I created with the assistance of GPT. We also listened and discussed an episode of Theories of Everything with Professor Jennifer Nagel.
Human Memory
In the episode, I slightly misremembered a study on memory related to the 9/11 terrorist attacks. Below are the main ideas related to my episode reference brought to you by GPT:
1. Memory decay slows after one year: Flashbulb memories (the personal context: where you were, who told you, etc.) and event memories (facts about the attack itself) both show a forgetting curve that flattens significantly after the first year.
2. Emotional details fade more than factual ones: Emotional responses associated with flashbulb memories (e.g., how shocked or upset you felt) are remembered less accurately over time than non-emotional details like location or the person who told you.
3. Subjective confidence remains high despite inaccuracies: By multiple follow‑up points (1 week, 11 months, 35 months), many participants recalled different details from their original reports—and yet they continued to express high confidence in their recollections.
After the initial decay, both flashbulb and event memories tend to become more stable—though not necessarily more accurate—over time.
Cumulative Selection
Cumulative selection can explain how a small advantage can build over time through natural selection. Ronald Fisher, a relatively unknown genius and polymath, developed mathematics that explain how small mutations that increase fitness, even with a very small advantage, can survive over time.
Using population genetics, a gene with a 0.1% selective advantage can dominate a population of 10,000 within 20,000 generation. That would be about 500,000 years for humans.
In the podcast episode, I generally explained why babies should look like their fathers. In theory, if babies, who looked like their fathers, had even a very small advantage, like 0.1%, that advantage would dominate human populations after 20,000 generation, or 500,000 years. If the advantage was larger, it would dominate faster.
Links