Mir(Mir)

Karma: 38

Flowers are selective about what kind of pollinator they attract. Diurnal flowers use diverse colours to stand out in a competition against their neighbours for visual salience. But flowers with nocturnal anthesis are generally white, as they aim only to outshine the night.

O mysterious temple of chance, for what did the scattered parts of your whole link arms to shape you? And the mysterious ceasefire they declared to heed you? O quiescent eye of a hurricane, what holds you together? O molecular weave of water and sunlight, why even bother?

Mir 20 Apr 2024 9:41 UTC
3 points
0
in reply to: niplav’s comment on: Self-Blinded L-Theanine RCT
Edit: I found the post usefwl, thankmuch!!
Mh, was gonna ask when you were taking it. I’m preparing to try it as a sleep-aid for when I adjust my polyphasic sleep-schedule (wanting to go fm 16h-cycles potentially down to 9h) bc it seems potentially drowsymaking and has much faster plasma decay-rate^[1] compared to alts. This is good for polyphasic if not want drowsy aft wake.

The data in ^[1] concerns 100mg tablets, however, and a larger dose (eg 400mg) may be longer. The kinetic model^[2] they use will prob be good estimate of plasma concentrations even if adjust dose.

Questions is whether it’s good estimation for duration of action in the brain, esp given that it’s “single-compartment model” (the blood is one compartment, and the brain is another). My heuristic for whether plasma T predicts brain T is whether the molecule v easily passes the BBB (as melatonin does), since then I can guess that the curve for the brain will look similar to the curve for the blood, offset slightly down and to the right.
1. ^
  Typical concentration-time curve of plasma ʟ-theanine of one participant after intake of 100 mg ʟ-theanine via one capsule (A) or 250 mL green tea (B). Circles represent measured concentrations of ʟ-theanine. The line represents the modeled plasma concentration-time curve by the use of the 1-compartment model.
  — Kinetics of ʟ-Theanine Uptake and Metabolism in Healthy Participants Are Comparable after Ingestion of ʟ-Theanine via Capsules and Green Tea, , 4 - ScienceDirect
2. ^
  Hot tip, you can screenshot pdf equations (i rec Text Grab or Greenshot) and ask gpt-4-turbo (updated) to write it in latex and/or jupyter-notebook-compatible python^[3].
3. ^
  Getting a tweakable python model for it was nontrivial after 35m of trying, so i’ll prob j wing it w 200mg 1h pre bedtime and do 3h sleep, and adjust bon feelings. ig the primary variable that determines how alert i feel upon waking up is getting the sleep-cycle timing right, and planning my wake-up routines (lights & text-to-speech-model as alarm) for j after I’ve done REM-sleep^[4].
4. ^
  While I probably would feel alert if waking up in middle of REM-sleep (after emerging from deep), I want to avoid that bc studies show bad effects from targeted deprivation of REM (leaving other phases untouched).

Mir 10 Mar 2024 14:25 UTC
−3 points
−10
on: Notes from a Prompt Factory
API requests should be automatically screened for human intent, and requests judged by the model to be disrespectfwl should be denied. (And they shouldn’t be trained to agree to respond to everything.)
I appreciate the post, but I also wish to hear more detailed and realistic scenarios of exactly how we might end up accidentally (or intentionally) sleepwalk into a moral catastrophe. I think it’s unlikely that punishment walls will make AIs more productive, but similar things may profitable/popular if advertised for human (sadist) entertainment.

Mir 31 Jan 2024 14:33 UTC
1 point
on: Predictive Categories Make Bad Causal Variables
this is rly good. summary of what i lurned:
- assume the ravens call a particular pattern iff it rains the next day.
  - iow, $\frac{P (raven's call | rain)}{P (raven's call | \neg rain)} = a bajillion$ , thus observing the raven’s call is strong evidence u ought to hv an umbrella rdy for tomorrow.
  - “raven’s call” is therefore a v good predictive var.
- but bribing the ravens to hush still might not hv any effect on whether it actually rains tomorrow.
  - it’s therefore a v bad causal var.
- it cud even be the case that, up until now, it never not rained unless the raven’s called, and intervening on the var cud still be fruitless if nobody’s ever done that bfr.
- for systems u hv 100% accurate & 100% complete predictive maps of, u may still hv a terrible causal map wrt what happens if u try to intervene in ways that take the state of the system out of the distribution u’v been mapping it in.
how then do u build good causal maps?
- ig u can still improve ur causal maps wo trial-and-error (empirically testing interventions) if u just do predictive mapping of the system, and u focus in on the predictive power of the simplest vars, and do trial-and-error in ur simulations. or smth.

Mir 9 Jan 2024 21:30 UTC
0 points
in reply to: Alexander Gietelink Oldenziel’s comment on: Johannes C. Mayer’s Shortform
u’r encouraged to write it!
You have permission to steal my work & clone my generating function. Liberate my vision from its original prison. Obsolescence is victory. I yearn to be surpassed. Don’t credit me if it’s more efficient or better aesthetics to not. Forget my name before letting it be dead weight.

Mir 8 Jan 2024 15:58 UTC
0 points
in reply to: mesaoptimizer’s comment on: Johannes C. Mayer’s Shortform
but u don’t know which distribution(s) u are acting in. u only have access to a sample dist, so u are going to underestimate the variance unless u ~Bessel-correct^[1] ur intuitions. and it matters which parts of the dists u tune ur sensors for: do u care more to abt sensitivity/specificity wrt the median cluster or sensitivity/specificity wrt the outliers?
ig sufficiently advanced vibes-based pattern-matching collapses to doing causal modelling, so my real-complaint is abt ppl whose vibe-sensors are under-dimensional.
1. ^
  idk the right math tricks to use are, i just wanted to mk the point that sample dists underestimate the variance of the true dists
  also, oops, fixed link. upvoted ur comment bc u complimented me for using RemNote, which shows good taste.

Mir 4 Jan 2024 2:45 UTC
0 points
in reply to: Johannes C. Mayer’s comment on: Johannes C. Mayer’s Shortform
Fwiw, you’re on my shortlist of researchers whose potential I’m most excited about. I don’t expect my judgment to matter to you (or maybe up to one jot), but I mention it just in case it helps defend against the self-doubt you experience as a result of doing things differently. : )
I don’t know many researchers that well, but I try to find the ones that are sufficiently unusual-in-a-specific-way to make me feel hopefwl about them. And the stuff you write here reflects exactly the unusualness what makes me hopefwl: You actually think inside your own head.
Also, wrt defending against negative social reinforcement signals, it may be sort of epistemically-irrational, but I reinterpret [people disagreeing with me] as positive evidence that I’m just far ahead of them (something I actually believe). Notice how, when a lot of people tell you you’re wrong, that is evidence for both [you are wrong] and [you are so much righter than them that they are unable to recognise how you are right (eg they lack the precursor concepts)].
Also, if you expect [competence at world-saving] to be normally (or lognormally) distributed, you should expect to find large gaps between the competence of the most competent people, simply because the tail flattens out the further out you go. In other words, P(you’re Δ more competent than avg) gets closer to P(you’re Δ+1 more competent than avg) as you increase Δ. This is one way to justify treating [other people not paying attention to you] as evidence for [you’re in a more advanced realm of conversation], but it’s far from the main consideration.
I invite you to meditate on this Mathematical Diagram I made! I believe that your behaviour (wrt the dimension of consequentialist world-saving) is so far to the right of this curve, that most of your peers will think your competence is far below them, unless they patiently have multiple conversations with you. That is, most people’s deference limit is far to the left your true competence.
I’m now going to further destroy the vibes of this comment by saying “poop!” If someone, in their head, notice themselves downvaluing the wisdom of what I previously wrote, merely based on the silly vibes, their cognition is out of whack and they need to see a mechanic. This seems to be a decent litmus test for whether ppl have actual sensors for evidence/gears, or whether they’re just doing (advanced) vibes-based pattern-matching. :P

Mir 4 Jan 2024 0:47 UTC
1 point
in reply to: Johannes C. Mayer’s comment on: Johannes C. Mayer’s Shortform
I notice that the mathematics-frame I used to try to generate a solution was utterly inadequate, whereas the programming-frame is much more productive wrt this problem. I think one big general weakness of my general math-frame, is that it imagines/visualises infinities as static, rather than as conceptually chunked dynamic processes.

Mir 3 Jan 2024 15:04 UTC
−1 points
−1
on: What Helped Me—Kale, Blood, CPAP, X-tiamine, Methylphenidate
Did you diagnose sleep apnea before getting the CPAP?
I’m surprised at your mention of kale (will look into it!) and pig’s blood. I think it’s good to mention here that pigs are prob sentient, and that it’s mean to eat them unless there are very strong utilitarian reasons to do so anyway. I think you prob have such reasons, because you are competently trying to save the world, so I’m glad you’re making moral sacrifices to get stuff done.
“Also, I now have a methylphenidate prescription which is pretty magical. I can now steer my mind.”
<3
Are you able to ask your psychiatrist for a larger dose? I think it could to at least 60mg, but idk.

Mir 2 Jan 2024 13:17 UTC
2 points
2
in reply to: quila’s comment on: quila’s Shortform
I don’t know the full original reasoning for why they introduced it, but one hope is that it marginally disentangles agreement from the main voting axis. People who were going to upvote based purely on agreement will now put their vote in the agreement axis instead (is the hope, anyway). Agreement-voting is socioepistemologically bad in general (except for in polls), so this seems good.

Mir 2 Jan 2024 12:54 UTC
−9 points
−2
in reply to: Dagon’s comment on: rationalistic probability(litterally just throwing shit out there)
if u can’t even tell if there are coherent ideas present, downvoting is socioepistemologically bad. imagine universalising this reaction across the website, and ask urself what happens to the ideas that are so novel that they don’t neatly fit into existing explanatory paradigms.
imo, u shud only downvote stuff u think is bad after u understand them. socially disincentivising illegible stuff seems quite bad in general.
if otoh u downvote bc it seems to neglect existing work in related fields, see the underappreciated value of original thinking below the frontier. we don’t want to disincentivise ppl from reinventing stuff. if they came up w schmobability theory on their own, i’d encourage them to expand on it, rather than risk collapsing the seed by making them learn abt Bayes.

Mir 30 Nov 2023 10:47 UTC
0 points
0
on: Stupid Question: Why am I getting consistently downvoted?
weird, i was intending that as a reply to @trevor’s answer, but it got plopped as its own answer instead.

Mir 30 Nov 2023 10:44 UTC
1 point
1
on: Stupid Question: Why am I getting consistently downvoted?
my brain is insufficiently flexible to be able to surrender to social-status-incentives without letting that affect my ability to optimise purely for my goal. the costs of compromise (++) btn diff optimisation criteria are steep, so i would encourage more ppl to rebel against prevailing social dynamics. it helps u think more clearly. it also mks u miserable, so u hv to balance it w concerns re motivation. altruism never promised to be easy. 🍵

Mir 9 Nov 2023 18:03 UTC
0 points
0
on: Steelmining via Analogy
Related recommendation: Inward and outward steelmanning — LessWrong
Imagine that you encountered a car with square wheels
Inward steelmanning: “This is an abomination! It doesn’t work! But maybe with round wheels it would be beautiful. Or maybe a different vehicle with square wheels could be beautiful.”
Outward steelmanning: “This is ugly! It doesn’t work! But maybe if I imagine a world where this car works, it will change my standards of beauty. Maybe I will gain some insight about this world that I’m missing.”
If you want to be charitable, why not grant your opponent an entire universe with its own set of rules?

Mir 9 Nov 2023 1:18 UTC
0 points
on: Procrastination Paradoxes: the Good, the Bad, and the Ugly
there has to be some point in time in which an agent acts like waiting just one more timestep before pressing wouldn’t be worth it even though it would.
if it’s impossible to choose “jst one mor timestep” wo logically implying that u mk the same decision in other timesteps (eg due to indifferentiable contexts), then it’s impossible to choose jst one mor timestep. optimal decision-mking also means recognising which opts u hv and which u don’t—otherwise u’r jst falling fr illusory choices.
which brings to mind the principle, “u nvr mk decisions, u only evr decide btn strats”. or fr the illiterate (:p):
You never make decisions, you only ever decide between strategies.

Mir 11 Oct 2023 11:26 UTC
1 point
0
on: Shard Therapy, preambe
It seems good to have an abstraction-level with separate concepts for soras and their weights.
I’m less enthusiastic about “sora” as the word for it, however, even though I like its aesthetics. Seems like it ought to be longer than 4 chars, and less isolated in meaning. I’ve wanted to find a more specific fit for the word “neureme” (from “emic unit”/”-eme”, categorising it as a unit of selection alongside gene, meme, morpheme, lexeme, grapheme, etc), and I like your concept for it.^[1] See here for my philosophy on word up-making.
1. ^
  This just means I’ll be using my word for it in my own notes, not that I’m recommending this relabelling for anyone else.

Mir 9 Oct 2023 12:09 UTC
1 point
0
in reply to: Johannes C. Mayer’s comment on: Write the Worst Post on LessWrong!
I was very very terrible at introspection just 2 years ago. Yet I did manage to learn how to make very good games, without really introspecting for years later about why the things work that I did.
More specifically, I mean progress wrt some long-term goal like AI alignment, altruism, factory farming, etc. Here, I think most ways of thinking about the problem are wildly off-target bc motivations get distorted by social incentives. Whereas goals in narrow games like “win at chess” or “solve a math problem” are less prone to this, so introspection is much less important.

Mir 9 Oct 2023 11:35 UTC
4 points
in reply to: Johannes C. Mayer’s comment on: adamzerner’s Shortform
I sort of deliberately created the beginnings of a tulpa-ish part of my brain during a long period of isolation in 2021 (Feb 7 to be exact), although I didn’t know the term “tulpa” then. I just figured it could be good to have an imaginary friend, so I gave her a name—”Maria”^[1]—and granted her (as part of the brain-convincing ritual) permanent co-ownership over a part of my cognition which she’s free to use for whatever whenever.
She still visits me at least once a week but she doesn’t have strong ability to speak unless I try to imagine it; and even then, sentences are usually short. The thing she most frequently communicates is the mood of being a sympathetic witness: she fully understands my story, and knows that I both must and will keep going—because up-giving is not a language she comprehends.
Hm, it would be most accurate to say that she takes on the role of a stoic chronicler—reflecting that I care less about eliciting awe or empathy, than I care that someone simply bears witness to my story.^[2]
1. ^
  Semi-inspired by hakomari, though I imagine her as much more mature in character & appearance than images I find online.
2. ^
  Oh yeah, and I’ve got the diagnostic triplet {ADHD, depression, aspergers (from back when that’s what it was called)} if that matters for anything.

Mir 7 Oct 2023 3:32 UTC
1 point
in reply to: Johannes C. Mayer’s comment on: Johannes C. Mayer’s Shortform
Oi! This was potentially usefwl for me to read.
WHEN I feel bad/uneasy at any point,
THEN find the part of my mind that’s complaining, and lend it my voice & mental-space.
I have previously tried to install a “somatic trigger” for whenever I feel bad (ie “when I feel bad, close my eyes and fold my hands together in front of me in a calm motion”), but it failed to take bc there weren’t clear-enough cues. The point of a somatic trigger in the first place is to install them in specific contexts such that I have clearer cues for whatever habits I may wish to write into those contexts.

Mir 7 Oct 2023 3:18 UTC
1 point
in reply to: Johannes C. Mayer’s comment on: adamzerner’s Shortform
Are those insights gleamable from the video itself for other people? And if so, would you be willing to share the link? (Feel free to skip; obviously a vulnerable topic.)

Mir 4 Oct 2023 18:57 UTC
0 points
0
in reply to: Johannes C. Mayer’s comment on: Computer Input Sucks—A Brain Dump
[Thoughts ↦ speech-code ↦ text-code] just seems like a convoluted/indirect learning-path. Speech has been optimised directly (although very gradually over thousands of years) to encode thoughts, whereas most orthographies are optimised to encode sounds. The symbols are optimised only via piggy-backing on the thoughts↦speech code—like training a language-model indirectly via NTP on [the output of an architecturally-different language-model trained via NTP on human text].
(In the conlang-orthography I aspire to make with AI-assistance, graphemes don’t try to represent sounds at all. So sort of like a logogram but much more modular & compact.)
When I write with pen and paper my writing improves in quality. And it seems this is because I am slower.
Interesting.
Anecdote: When I think to myself without writing at all (eg shower, walking, waiting, lying in bed), I tend to make deeper progress on isolated idea-clusters. Whereas when I use my knowledge-network (RemNote), I often find more spontaneous+insightfwl connections between remote idea-clusters (eg evo bio, AI, economics). This is bc when I write a quick note into RemNote, I heavily prioritise finding the right tags & portalling it into related concepts. Often I simply spam related concepts at the top like this:
The links are to concepts I’ve already spotted metaphors / use-cases for, or I have a hunch that one might be there. It prompts me to either review or flesh out the connections next time I visit the note.
I think these styles complement each other very well.