January 3rd, 2007



Counting exercise again, a bit less than 10 minutes this time.

Started off thinking about possible new features for our software at work. Followed that with a little bit of thinking about darkone238's and my relationship. Spent a bit more time after that remembering how disappointed I was in my mom's behavior at my graduation ceremony. Then I started getting drifty again--moreso than last night--leading me to the conclusion that I need to be getting more sleep. I didn't lose the count at all, though I came close to doing so on a couple of occasions due to drifty moments.

My breath felt a little less constrained/stressed this time. It was still heavy, and maybe a bit more ragged, but more evenly-spaced. My back was mildly sore, but generally so -- the uncomfortable pain in my mid-back from last night didn't develop. I started (and ended) the session with a headache which covered my left temple and eye. My left eye was twitching a bit at the beginning, but relaxed as the session progressed. I found that when I focused more attention on it, it twitched more.


I'm trying to decide if splitting the entry into two paragraphs of general impressions over time is the right thing to do. I'm wondering if a better approach might be to stick to a strict chronological sequence of events (interleaving mental and physical as they occur), at least insofar as I can remember them.

Telling my brain to remember things for later seems to be highly inefficient -- at least, it requires a lot of CPU time. That's not surprising, though. My brain does much better at remembering interrlelated concepts with clearly-defined relationships than it does a strict sequence of facts with no defined semantic relationships between them. It's as if it has a built-in compiler that takes such facts and immediately reduces them to a system (or model), then iterates over the system, repeatedly reducing it further, thus making it more general while still mostly preserving its correctness.

That's how I imagine my brain works, anyway. In practice I'm sure it has many different representations for different tasks. It's like the opposite of programming -- taking undifferentiated data and using it to operate on a program, rather than the other way around.

Hmm. That's something interesting to think about. I'll have to play with that idea a bit more -- I wonder if it can somehow be refined or translated into terms that are more useful for writing "smart" programs (e.g. spam filters and the like).

-- Des