The forgetting curve appears roughly exponential.

Let's hypothesize further: Perhaps the forgetting curve is some kind of Poisson distribution-related thing.

A "train of thought" is a stochastic process by which you jump from thought A to thought B to thought C. A thought is a gestalt phenomenon of many different neurons firing at once; then there are a great deal many possible thoughts that could slot into B depending on which ->

<- neurons actually fire as a result of A.

But to take any *specific instance* of thought B, either B will be that specific instance exactly or it won't, and the chance that it will is staggeringly low.

That's a Bernoulli random variable!

This suggests a competing process in the brains of ADHD types, like my friend Mason and my brother Alan. On the one hand, their brains fire off trains of thoughts extremely quickly - this is a buffering effect against forgetting for any thoughts that happen to be triggered ->

<- but this effect is swamped by the sheer stochasticity of the process. The specific gestalt of thought B they *want* to not forget may indeed become more and more likely over time as they wander down the same memory trace, but there are just too many possibilities

i have very little clue whether this will make sense to anyone who isn't myself so i apologize. my own neuron gestalts are way more xNFx than xNTx and i refuse to change