Random notes.txt

https://youtu.be/-RdOwhmqP5s?t=832

Chaos at the boundaries between regions, where each region represents contiguous points that end up having the same root as the root they are closest to after N iterations.

P vs NP, where many NP-complete problems are easy in the region where the answer is obviously yes or obviously no, but hard in the region where the answer approaches the boundary.

https://youtu.be/WFbgfknnWqA?t=308

S[x_i(t), x_i'(t)] = Integ[min=t0,max=t1] L(x_i(t), x_i'(t)) dt

S is not a zero-order function; it is a "functional".

A functional is a higher-order function from zero-order functions to numbers.

In this particular case, the x_i(t) functions are functions of the same variable, namely time.
I think they were velocity and acceleration within each spatial dimension?
Anyway, the x_i(t) functions and their d/dt derivatives ARE the path under consideration.

Minimizing S[...] means minimizing dS/dx_i for (each? just one?) function x_i(t).

This means minimizing the degree to which small differences in x_i(t)'s value on the [min=t0,max=t1] integration range can change L.  (Chaos attractor!)

L = T - V, for reasons that don't make sense classically.
L = Lagrangian
T = Kinetic Energy
V = Potential Energy

Video asks: how do particles "know" which path to take to minimize action, if action implies knowing the start and end states?

QM action is all about probability
QM allows multiple consistent paths, unlike CM
Double slit extension: more slits, more amplitudes that we have to add to predict the pattern on the screen; in limit of infinite slits (no barrier) we get "normal"
Feynman Path Integral: all paths contribute.  Particle doesn't need to know the minimized action path ahead of time, it simply does all paths at once.
CM action IS QM amplitude: Lagrangian function determines the quantum amplitude for each possible path.
Paths far away from most likely paths tend to curl their amplitudes around each other in R^2 aka C^1 vector-space, thus their contributions average to zero.
Paths close to the most likely paths tend to point their amplitudes in a similar direction, thus their contributions average to non-zero.
The CM least action path is the point where the QM paths have the most coherent directions for their amplitudes.
Increasing 1/hbar increases the rate at which the amplitude rotates,
and the faster the amplitude rotates the fewer paths exist that have a coherent amplitude direction with one another,
and therefore the fewer the number of QM paths that the particle might actually have traveled,
and therefore the less likely a violation is of the rule that the CM least action path always wins.
At hbar = 0, CM is exactly recovered.
And because QM amplitude is CM action, the CM "least action" path...

BREAK

https://youtu.be/qtf6U3FfDNQ?t=50
Symmetries are transformations that can be applied to a physical system that leave the physical system in the state in which it began.  (indistinguishable)
What if we're thinking backwards by thinking that every transformation is possible but symmetries make some of them null?
What if transformations are not possible unless they change the state?
Possibly tangentally related to discrete vs continuous in relation to my belief that there are no physical infinities or infinitesimals.

BREAK

thinking about "evil" vs thinking about "consequences"
w.r.t Ana Mardoll/Lockheed-Martin

BREAK

Uncertainty Principle viz energy-time
The more the energy of the system is allowed to fluctuate, the shorter these fluctuations can last.
A system can borrow energy
as long as it quickly gives the energy back.
The Big Bang = tiny amount of energy that produces a single lightweight particle ("the creaton") that somehow triggers a cascade of pair production AT ZERO ENERGY
That pair production is still happening today, the creaton still exists in modern times
Because the universe itself is still net positive energy UNTIL it pays the energy back at the end of the heat death and the creaton fades away
So either pair-production NEVER creates creatons (sounds bogus) OR the cascade of pair production is somehow suppressed in the sea of dark creatons that surround you
The creaton interacts with some new field mediating the "creaton charge"
The "creaton charge" vs "energy scale" graph for the creaton is mostly flat, but suddenly blows up at a very high energy scale
Aside: Maybe it's shaped like O(2**poly(E))
The creaton is like a bomb that doesn't stop exploding, it's t=0 for each of the inflationary multiverses that are constantly budding off from it
The bud-offs happen when the chaos finally cools down
We are a metastable state within that cauldron of creation

BREAK

Mach bucket experiment
lim as gravity goes to 0: bucket emptied as water flings everywhere (surface no longer parabolic)
lim as gravity goes to infinity: bucket surface level remains flat
absolute angular momentum comes from infinite "background" gravity that applies equally in all directions?

BREAK

Orbital "Monsters Exist" Is Good, Actually
Best since Snivilization
P.E.T.R.O.L.	drums in first minute of build-up
Lush 3-1	riff after build-up
Halcyon		riff after build-up (same as Lush)

BREAK

Series of blog posts: intro to media literacy, wanna do thematic analysis of
Tarantino's Reservoir Dogs just b/c Tarantino is one of the least theme-heavy
directors out there

AI Post 3: community of mind, brain modules voting on what's important right now,
the morals populating our actual Spreadsheetist ethical system are each championed
by exactly one brain module, which provides a clear pathway to explaining how
multiple agents can aggregate their moral preferences to create a shared ethical
framework that makes both people happy, assuming neither person's morals directly
conflict with the other's

EY fanfic:
1. PTSD from autism and ADHD as evidenced from EY's fiction writing.
2. Breakdown of how "having insights quickly" is not intelligence, it's ADHD and/or a trauma response.
3. Insight depends on background knowledge plus analogy / synaesthesia and has nothing to do with intelligence.

BREAK

P vs NP

What shape is the space of 3SAT instances?
What does a fractal look like in a discrete space?
Are hard instances of NP-complete problems hard because they lie in regions of the instance-space where arbitrarily small deltas in the instance choice lead to differences in which attractor (True or False) the computation ends up at?

P vs BQP

Given a decision problem, what is the relationship between the minimum circuit depth and the minimum reversible circuit depth for that problem?
My understanding is that reversible circuits are complete but that irreversible circuits offer a speed-up / depth reduction at the cost of increasing entropy elsewhere.
Given that, how could there possibly exist problems whose min depth reversible circuits are deeper than their min depth irreversible circuits?  Isn't that a requirement if P < BQP?

BREAK

moral obligation to continuously learn
world is inherently dangerous, we create safety, that doesn't stop being true when we reach adulthood
Sealed ampule of cesium as a shiny toy on Reddit
Goiânia accident
my psychiatrist, outdated neuroscience
serotonin imbalance hypothesis is proven wrong
new neurons born each night when you sleep, they migrate into place over the next 2 weeks and form synapses
my guess: almost all synapse remodeling occurs during cell infancy, synapse remodeling in adult cells indicates distress
my guess: schizophrenia is caused by over-eager synapse formation not balanced by pruning, too many irrelevant associations
tangent over
I need to know neuroscience and moral philosophy *because* I'm a computer programmer and need to be responsible with AI
sentience is more important than consciousness from a moral perspective
sentience is layer 1
moral agency is layer 2
consciousness is layer n, for some n ≥ 3
consciousness without sentience does not need to be treated morally, though
moral agency is not possible without sentience; it is sentience + awareness that other moral agents exist
consciousness without sentience, if it is possible, does not care about its own continued existence and therefore neither should you
consciousness is That-Which-Speaks
TWS is a brain component has the job of observing other brain components, forming hypotheses about them, and offering high-level suggestions
TWS does not command the brain, but it leads the brain
the other brain components follow TWS's leadership only if they agree with it
TWS is the only part of the brain with language
other brain components communicate with TWS using feelings, associations, memories
TWS is tasked with speech and writing, but that probably came later
TWS probably exists in all vertebrates, at least in some form
TWS has observable behavioral consequences: have you ever seen a cat that was about to do something, but stops and freezes for ten seconds, then gets up and does something else?  That's TWS