I know one concept of String Theory is that every time you make a choice, a new parallel universe is created. I am not sure if Tom agrees with this idea.
another version of what Ted said...
rather, every time you make a choice, what was previously one of an array of probabilities, becomes an actualized choice, and this births a new array of plausibilities and probabilities, each with levels of significance, from which conciousness must next choose, with the computer working ahead to pre-calculate what could happen at the DELTA(t) after the next decision, and on and on up to a certain point of unprofitability
so each decision creates this new web or fan of possibilities just ahead of you, and leaves a narrow trail of actualized decision, within a spiderweb of what could have happened but didn't. For system learning, this morass of what happened and didn't happen is what is useful - what actually happened may not actually be more important than what could have happened, Tom spoke to that on the Boston record if I recall in response to a question I put on this - so I presume, what actually happens is merely like a trailer bullet or scout to guide the overall process, from the POV of AUM or simulation management.
for the individual, it can be sort of spooky to be fully aware of these decision tree forks in the road in real time, realizing that a choice, turning this way or that way at a corner, will construct an entirely different life for you. The guidance is to increasingly dial up your concern for "other" at each of these forks in the road, and this leads to a much more comfortable and fun life, somewhat ironically.