Arguably both, since I thought of it but later found I've been beaten to it at least twice. Steven Lehar beat me by 20 years and made a phd thesis out of it.
To give a quick summary of what I'm talking about, the brain waves rather than the individual neurons are the representation of everything. Brain waves are a consequence of closed loop neuron paths and propagation delay. The longer the loop, the lower the frequency. One consequence of this is that complex signal processing tricks like fourier transform fall out very easily, far more easily than if you tried to train a regular neural net to do it. Now if you poke at the timings or connections at random, you are more likely to hit a long loop than a short loop. Hence, LSD distorts a very particular band of brain wave frequencies. You can make words just fine but sentences lose their way.
http://slehar.com/wwwRel/
Psychedelic Information Theory also echos basically everything I was conjecturing at the time and backs it up in a ton of depth.
http://psychedelic-information-theory.com/
To give a quick summary of what I'm talking about, the brain waves rather than the individual neurons are the representation of everything. Brain waves are a consequence of closed loop neuron paths and propagation delay. The longer the loop, the lower the frequency. One consequence of this is that complex signal processing tricks like fourier transform fall out very easily, far more easily than if you tried to train a regular neural net to do it. Now if you poke at the timings or connections at random, you are more likely to hit a long loop than a short loop. Hence, LSD distorts a very particular band of brain wave frequencies. You can make words just fine but sentences lose their way.
A paper that confirms the fourier trick is going on in the optical system: http://www.math.utah.edu/~bresslof/publications/01-3.pdf
Experimental evidence of the shift in frequency partition function due to LSD https://www.nature.com/articles/s41598-017-17546-0.