I want to hear gwern’s research into the Algernon problem for it, first.
edit: Since both gwern and wedrifed thought I was asking for reassurance—I want to be reasonably confident that there isn’t an as-yet unnoticed Algernon problem with this entirely new type of brain hacking before using it on myself.
I have none so far. Without a mechanism of action, I can’t assess what criterion it might satisfy to avoid Algernon’s law. The speculation about it upregulating NMDA to change plasticity is not enough to go on—at this point I can only offer the generic observations that there are probably metabolic limits to learning in the ancestral environment but they would not apply to ours and this may be why our learning is regulated below its peak.
EDIT: For example, there’s a whole theory of sleep (“synaptic homeostasis”) which says it exists solely to reduce overall synaptic potentiation: the more learning you do, the more synapses fire more, and the more energy the brain needs just to function. The effect isn’t small: something like >10% by nightime in one study the author cited.
Since both gwern and wedrifed thought I was asking for reassurance
No I didn’t—and I gave neither reassurance nor the the reverse. I questioned and attempted to correct emphasis on finding Algernon based ‘research’. We don’t have anything of the sort when it comes to specific techniques like this, such research would be difficult to the point of implausibility and not of much practical benefit if we had it. It would satisfy our historic curiosity but little more. Research on the actual effects of the technique on humans screens off questions of why our brains don’t have equivalent methods in place.
My apologies for misreading. Yes, Algernon’s Law motivates the research (in addition to “my brain is important and this is weird,” and other motivations); but research shouldn’t be focused on whether the mechanisms were available in the EEA unless that evidential node is among the low-hanging fruit, as far as ease of research and large likelihood ratios go.
I want to hear gwern’s research into the Algernon problem for it, first.
I hypotheses about why things avoid the Algernon problem generalize and attributing a specific technique or tool to any given Algernon bipass is largely a matter of telling just so stories. Far, far more interesting (or failing that more useful) is what research gwern can find whether this stuff works and is safe. We don’t need him to find an excuse for it to work and he doesn’t have enough information available to make strong credible claims about Algernon’s law as applied to TDCS.
I want to hear gwern’s research into the Algernon problem for it, first.
edit: Since both gwern and wedrifed thought I was asking for reassurance—I want to be reasonably confident that there isn’t an as-yet unnoticed Algernon problem with this entirely new type of brain hacking before using it on myself.
I have none so far. Without a mechanism of action, I can’t assess what criterion it might satisfy to avoid Algernon’s law. The speculation about it upregulating NMDA to change plasticity is not enough to go on—at this point I can only offer the generic observations that there are probably metabolic limits to learning in the ancestral environment but they would not apply to ours and this may be why our learning is regulated below its peak.
EDIT: For example, there’s a whole theory of sleep (“synaptic homeostasis”) which says it exists solely to reduce overall synaptic potentiation: the more learning you do, the more synapses fire more, and the more energy the brain needs just to function. The effect isn’t small: something like >10% by nightime in one study the author cited.
No I didn’t—and I gave neither reassurance nor the the reverse. I questioned and attempted to correct emphasis on finding Algernon based ‘research’. We don’t have anything of the sort when it comes to specific techniques like this, such research would be difficult to the point of implausibility and not of much practical benefit if we had it. It would satisfy our historic curiosity but little more. Research on the actual effects of the technique on humans screens off questions of why our brains don’t have equivalent methods in place.
My apologies for misreading. Yes, Algernon’s Law motivates the research (in addition to “my brain is important and this is weird,” and other motivations); but research shouldn’t be focused on whether the mechanisms were available in the EEA unless that evidential node is among the low-hanging fruit, as far as ease of research and large likelihood ratios go.
I hypotheses about why things avoid the Algernon problem generalize and attributing a specific technique or tool to any given Algernon bipass is largely a matter of telling just so stories. Far, far more interesting (or failing that more useful) is what research gwern can find whether this stuff works and is safe. We don’t need him to find an excuse for it to work and he doesn’t have enough information available to make strong credible claims about Algernon’s law as applied to TDCS.