Creativity is another big area that seems neglected. I’ve read a fair amount on the subject, but feel I have barely touched the surface.
I also feel it probably is relevant to AI, so I’m somewhat surprised to see so little discussion of it here. (By “AI” I mean a number of things here. Might be easiest to see it as application of computers to solve problems.) At the moment, AI works when the actions one can take are clear (e.g., small number of valid moves in a game). When the possible actions are not precisely specified, the specification becomes the issue. Generating these possibilities is not trivial, and frequently this is what creativity is.
You asked for “S1 training.” S1 means “System 1″ right? System 1 training is done by reinforcement, and trigger action planning is the mechanism by which one optimally sets and resets such reinforcements.
So if that’s not what your want, then what are you asking for?
Care to explain? Reinforcement training relies on providing the right feedback at the right moment for maximal effect. A trigger action plan is how you set an “alarm” in advance to arrange for maximal impact. This enables optimal reinforcement per unit of effort put in (or if not, then that is indicative that a better trigger action plan could have been used).
You made a rather strong claim, I expressed skepticism. Reinforcement loops do play in to deliberate practice (which is the strongest model of how experts attain expertise AFAIK), but jumping from deliberate practice to taps is a non-obvious transformation.
S1 training. I.e. creativity training, physical training, hedonic resetting, internal trust building etc.
Creativity is another big area that seems neglected. I’ve read a fair amount on the subject, but feel I have barely touched the surface.
I also feel it probably is relevant to AI, so I’m somewhat surprised to see so little discussion of it here. (By “AI” I mean a number of things here. Might be easiest to see it as application of computers to solve problems.) At the moment, AI works when the actions one can take are clear (e.g., small number of valid moves in a game). When the possible actions are not precisely specified, the specification becomes the issue. Generating these possibilities is not trivial, and frequently this is what creativity is.
Search term is “trigger action planning”.
That is a very different beast from deliberate practice of feldenkrais for example.
You asked for “S1 training.” S1 means “System 1″ right? System 1 training is done by reinforcement, and trigger action planning is the mechanism by which one optimally sets and resets such reinforcements.
So if that’s not what your want, then what are you asking for?
I’m not sure this adequately describes S1 training.
also not sure optimal is true.
Care to explain? Reinforcement training relies on providing the right feedback at the right moment for maximal effect. A trigger action plan is how you set an “alarm” in advance to arrange for maximal impact. This enables optimal reinforcement per unit of effort put in (or if not, then that is indicative that a better trigger action plan could have been used).
You made a rather strong claim, I expressed skepticism. Reinforcement loops do play in to deliberate practice (which is the strongest model of how experts attain expertise AFAIK), but jumping from deliberate practice to taps is a non-obvious transformation.