unprompted/Notes/art-of-doing-science.md

Rough Notes

Questions

Are people more alike, or different than machines?

• Since I'm just talking to myself here, I'll pin that as a similarity, more than a difference

  • Machines (I'm specifically thinking of computers) spend a great deal of time generating and handling internal messages and signals.
  • It's worth reiterating that networked communication came later, and before that time, the bulk of external communication was with a single user through some narrow interface

• Machine's don't belabor the unoriginality of their output. They just produce prodigious amounts of structured data. A great deal of it mind-numbingly repetetive.

  • So here I'd say that more effective people, less concerned or completely indifferent to the novelty of their thinking,
  • but rather focused on the appropriateness, given the context, or the richness of taking an idea off the shelf and applying it in some new way.
    • "The Terroir"
  • Less effective people are constipated, preoccupied with the wrong evaluation.
  • Ideas get propagated because they're reactive
  • You can have truly original ideas that are deranged and go nowhere (maybe they just rattle inside your head.)
  • But the ideas that get repeated, passed on, have a "contageous" quality that doesn't burn out
    • They're enticing. Why?
      • A lot of marketing - you'll like what you're told to like
        • But that starts to feel circular:
          • like what you're told to like
          • told to like what people like
        • We might try on an idea, walk it around a bit
          • But I don't think marketing pressure alone can sustain it
          • Needs some dose of intrinsic appeal
            • But even that acclimation can come from repeated exposure.
            • So again, the things we want are the things we've been routinely exposed to by one factor or another
      • Some intrinsic value (or potential - does that even exist?)

• Some machines are serviceable, others are not. In general, people are only serviceable by other people with specialized knowledge and training. Same goes for a lot of integrated circuits I've met, CPU dies and the like.

  • So this isn't a great defining quality for machines. It's a better one for people
  • A better way of phrasing it, to avoid trending into irrelevance:
    • some machines exhibit modularity. Parts can be swapped out for compatable ones.
      • But then, that sounds closer to people. Maybe we're not hot-swappable, but we face compatability issues. Blood types, organ donations. Could maybe stretch it to overlap with the parts metaphor, where maybe a badly compatible part could be swapped in disasterously.

• So the similarities and differences have more to do with an attitude than an absolute.

  • I wouldn't know how to prove that there will always be counter-examples, or alternative perspectives for every argument made for one camp or the other.

Quotes

Definitions

Stirling approximation of n!

• `ln(n!) ~= n * ln(n) - n
• or `n! ~ sqrt(2pin) * (n/e)^n
  • why does a circle appear?
• logarithm rule

1. let a = e^p s.t. ln(a) = p

2. let b = e^q -> ln(b) = q

3. a * b = e^p * e^q = e^(p + q)

4. ln(a * b) = p + q = ln(a) + ln(b)

5. qed motherfuckers

• ln(1) = 0

Power Rule

(u * v)' = u'v + uv'

(x^2 * sin(x))' = 2x * sin(x) + x^2 * cos(x)

Integration by parts

int((uv)') = uv = int(u' * v dx) + int(u * v' dx)

Thinking

• WTF is thinking, anyway?