__nobody

Karma: 249

Nobody’s Conversation Menu

(some of these are things that I thought & read about a lot already, others are more of a “I should do more here” thing… talk to me to find out which ones are which)

DMs here should work, you can also reach me on Discord as nobody2342 or on Telegram as @locally_cartesian. (Or ask for Matrix or Signal.)

Programming & Computations

dependently typed programming & formal software verification
non-standard logics (mostly constructive so far, but looking into more)
LANGSEC
how types, programming and exploitation relate to each other
metaheuristics / old-fashioned optimisation approaches
alternative approaches to software versioning, towards (non-containerised) software still usable in 100+ years

I should also learn about the current de-facto standard way of writing software (docker, JS / Typescript, …) at some point…

Community

According to wiki.hackerspaces.org Berlin has 18 hackspaces (tho a good chunk seems inactive / planned only), I’ve been to ¹⁄₃ of them and plan to visit the rest. I’ve also regularly been in hackspaces for over 15 years.

Planning & Knowledge Management

I’m basically running a hackspace at the moment. Don’t really know a lot about planning etc., so throw valuable info my way!
At some point I was using 5 tools to keep track of my plans and todos, one of them I made myself / am still working on. (The other tools are TaskWarrior, Complice, Obsidian and a calendar.) Currently I’m undecided between making a single “better” thing, or just waiting for LLMs to become usable locally with less knowledge / on worse hardware, and then let that figure out connections.

Food

things I cooked last week & what I should cook next week
ways to make tasty food quickly, or with very low effort
fermentation

Well-being & health

dealing with chronic pain / body malfunctions
relaxation / releasing tension
autism problems (being overwhelmed by your surroundings and how to deal with that)

Weird Brain Philosophy Stuff

What does reality feel like for you?
What are the axioms of your experience?
How do you shift your perception of reality?

Potential Future Interests

Should I care more about EA?

__nobody Oct 1, 2021, 2:33 AM
6 points
in reply to: elspood’s comment on: The 2021 Less Wrong Darwin Game
I didn’t trusty myself to reimplement the simulator—any subtle change would likely have invalidated all results. So simulations were real slow… I still somehow went through about 0.1% of the search space (25K of about 27M possible different species), and I hope it was the better part of the space / largely excluding “obviously” bad ideas. (Carefully tweaking random generation to bias it towards preferring saner choices, while never making weird things too unlikely.) Of course, the pairings matter a lot so I’m not at all certain that I didn’t accidentally discard the best ones just because they repeatedly ended up in very unfortunate circumstances.

There certainly were some kinda non-intuitive choices found, for example: A Benthic creature that can (also) eat grass—it can’t start in the river, but that’s where it wants to go; and travel-wise, Ocean/Benthic are equivalent! (Also, for some reason, others trying the same strategy in the ocean performed way worse… absolutely no idea why yet.)

I’d have loved for this to happen in a less-busy week (not exactly the end of the quarter year with all the bookkeeping) and to have about 2-3x as much time to get the infrastructure working… managed to barely get simple mutation working, but didn’t have time for the full genetic algorithm or other fancy stuff. :(

__nobody Sep 29, 2021, 9:07 PM
1 point
in reply to: elspood’s comment on: The 2021 Less Wrong Darwin Game
Seconding this, does ‘by Sep 30th’ mean start or end of the day? I’m currently assuming ‘end of’, in some unspecified time zone.

My computer’s still crunching numbers and I’m about to head to bed… would be sad to miss the deadline.

__nobody Sep 25, 2021, 3:54 PM
1 point
in reply to: Measure’s comment on: The 2021 Less Wrong Darwin Game
I’ve got it mostly working now… problem is that the default plot size is unusable, the legend overlaps, etc. etc. -- when run interactively, you can just resize the window and it’ll redraw, and then you save it once you’re happy. So now I’m also setting plot size, font size, legend position, and then it’s “just” a (plt.savefig "plot.png") added before the (plt.show).

I might also add a more easily parseable log output, but for now tee and a small Lua script mangling it into CSV is enough.

I’ll probably clean up all of that in a couple of hours and send another patch / merge request. Tho it’s probably a good idea to get someone else to review it given that I have near-zero prior experience w/ Python.

__nobody Sep 25, 2021, 2:14 PM
1 point
on: The 2021 Less Wrong Darwin Game
Another question: where does the script save the data / graphs when running it? Or does it do that at all?

It looks like it might try to open a plot window, but I’m running it on a headless server… so nothing will happen. Is the (hard-to-parse) text scrolling by all that I’ll get at the end of a run?

__nobody Sep 25, 2021, 1:47 AM
3 points
in reply to: lsusr’s comment on: The 2021 Less Wrong Darwin Game
Thanks!

Turns out it needs a newer hy, then it works. (And in case anyone else has a similar problem and is also a Python noob, pip’s package is probably called python3-pip or something like that. After that, the rest is explained either in the article or by pip itself.)

__nobody Sep 25, 2021, 1:05 AM

1 point

in reply to: lsusr’s comment on: The 2021 Less Wrong Darwin Game

That just gets me an even longer error message:

Python 3.7.3 (default, Jan 22 2021, 20:04:44)
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import hy
>>> import main
Traceback (most recent call last):
  File "/usr/lib/python3/dist-packages/hy/macros.py", line 238, in reader_macroexpand
    reader_macro = _hy_reader[None][char]
KeyError: '*'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<frozen importlib._bootstrap>", line 983, in _find_and_load
  File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 668, in _load_unlocked
  File "<frozen importlib._bootstrap>", line 638, in _load_backward_compatible
  File "/usr/lib/python3/dist-packages/hy/importer.py", line 197, in load_module
    self.path)
[...(*snip*)...]
  File "/usr/lib/python3/dist-packages/hy/compiler.py", line 424, in compile_atom
    ret = _compile_table[atom_type](self, atom)
  File "/usr/lib/python3/dist-packages/hy/compiler.py", line 365, in checker
    return fn(self, expression)
  File "/usr/lib/python3/dist-packages/hy/compiler.py", line 2559, in compile_dispatch_reader_macro
    expr = reader_macroexpand(str_char, expression.pop(0), self)
  File "/usr/lib/python3/dist-packages/hy/macros.py", line 242, in reader_macroexpand
    "`{0}' is not a defined reader macro.".format(char)
hy.errors.HyTypeError:   File "/home/nobody/projects/darwin/main.hy", line 63, column 1
[and then as before]

Guess I’ll try setting things up on another system tomorrow, or else I’ll skip.

__nobody Sep 25, 2021, 12:51 AM

1 point

on: The 2021 Less Wrong Darwin Game

Do you have a more exact version spec? Because I don’t even have pip3 and I don’t use Python, so I just installed the hy that comes with my distro… and then I get

  File "./main.hy", line 63, column 1

  (defn initial-population [biome]
    "Creates a list of initial organisms"
    (+ [] #*
       (lfor species +species-list+
             (if (= (. species ["spawning-zone"])
                    biome)
               (do
                 (setv organisms [])
                 (setv energy +seed-population-energy+)
                 (while (> energy 0)
                   (.append organisms (.copy species))
                   (-= energy (energy-of species)))
                 organisms)
               []))))
HyTypeError: b"`*' is not a defined reader macro."

and I don’t even have a clue what the b"``*' part is supposed to mean...

I experimented some more, no clue what package I’d have to install to get pip - neither pip nor pip3 exist. On the systems that I have, I get matplotlib 3.0.2-2 with hy 0.12.1-2.

__nobody Sep 9, 2021, 2:10 AM
6 points
on: LessWrong is providing feedback and proofreading on drafts as a service
Once things have stabilized and things like inline annotations are there, I’d love to see the following: (1) An easy way to add and remove yourself from a pool of available feedback providers. (Checkbox in settings?) And (2) a way for anyone (or nearly anyone—e.g. non-negative karma) to request brief / “basic feedback” on their posts, by automatically matching people from the pool to posts based on e.g. post tags and front page tag weights.

On (1): I have proofread a couple thousand pages by now, and while I’m usually pretty busy, in a slow week I’d be happy to proof-read a bunch of drafts. However, if that involves messaging someone to be added to a list and then messaging someone again to be taken off that list, that’s quite a lot of overhead—I probably wouldn’t bother and just look for other stuff to do. So I suspect automating that part might greatly increase the amount of available reviewers for feedback.

On (2): With some extra capacity from (1), I expect the main bottleneck for providing more reviews is matching reviewers to posts. If that’s automated, the only cost is the time spent reviewing. With a scheme like “review two/three posts to get one post reviewed by two/three people”, the reviewer pool should be even bigger (so with (1) you might even go “review 2 to get 3″) and things should remain relatively fair. With multiple reviews, you should get some decent feedback even if one reviewer writes complete nonsense or doesn’t understand anything.

At that point, the human overhead for having this extra “basic feedback” system should be near-zero, apart from from maybe having to manually filter people trying to abuse the system—no clue how prevalent that is. And looking at myself, (a) I probably wouldn’t bother manually asking others for reviews, and (b) knowing that I can get guaranteed feedback, no questions asked, would make it more likely to actually start writing. (While I can’t say for sure whether that translates into actual posts, I can clearly see that there are lots of other “very important” things, some of which only barely win out because they’re less headache-inducing / uncertain.)

__nobody Dec 21, 2020, 12:00 PM
1 point
in reply to: lejuletre’s comment on: Gauging the conscious experience of LessWrong

Could you imagine the feeling of lying on a carpet without a shirt on (ie the feeling of a carpet on your torso) ?

Somewhat… it’s too diffuse. I can imagine the effect at single spots, the whole thing at once doesn’t really work. (I get “glitchy partials”, brief impressions flickering and jumping around, but it’s not forming anything consistent / stable.)

What about a spider crawling across your hand ?

Back of the hand is manageable (it’s “only” tracking of 9 points − 8 legs plus occasional abdomen contact) and it can even become “independent” and surprise me with what direction it will move in next, front is basically impossible. (Glitchy partials again.)

I am very jealous of your ability to ignore your thoughts[.]

If there’s usually not much happening that reaches the conscious level, that’s really not that hard. I’ve talked at length with people who have a near-constant internal monologue and I get that that’s much harder. I just notice that it’s also relatively easy for me to ignore other constant or near-constant things like loud buzzing noise (server fans?), most extreme smells (rotting/feces) where others flinch away, etc. but basically impossible to ignore constantly changing impressions. (I currently have a construction site in front of the house and I can’t work at all while they’re active, even with earplugs.)

[I am very jealous of your ability to] track north.

Again, I have to constantly keep an “inner eye” on that or it comes loose and then it’s broken / de-synced. A question from someone that makes me think just a little bit too much can be enough to break it, unless I remember to explicitly save the current heading and reinstate the thing afterwards. (I suspect you can do/imagine something similar and it’ll get less costly and more precise over time of using it.)

I wonder if there’s any tangible overlap between brain function and fields of interest.

Probably—you’re more likely to do stuff that’s easier for you, which makes it easier… But I expect that to be fairly weak / to have lots of noise on top. (External expectations, stuff that you don’t even know exists, etc. etc.)

I guess what also plays a big role is how you approach your brain / how you model it & yourself / what expectations you place on it. I treat mine as a substrate that can spawn lots of independent processes of varying size and with more or less expressive interfaces, and “I” just happen to be one that’s fairly big and stable. So making that arrow pointing north is basically just me spawning yet another small process on that substrate that subsequently does its own thing (unless swapped out because of capacity constraints), likewise the imagined spider can become independent and choose its own direction, because I fully expect it to be an independent process and not something that I have to manually control.

(Treating your brain as some sort of “expectation realizer” seems to be a powerful model/perspective, but totally expect really weird “replies” when you try to apply that to external stuff. (Like weird feelings, sudden anxiety, strange preferences, etc. guiding you in the direction that your brain thinks is best—be careful (and keep track of) what you wish for.) Internal seems to be relatively safe compared to that.)

__nobody Dec 20, 2020, 6:34 PM
3 points
in reply to: wild-green’s comment on: Gauging the conscious experience of LessWrong

Also one of my friends struggles with verbal thinking and thinks mostly implicitly, using concepts, if I understood that correctly, and they have a strong preference for non-verbal signs of affection (physical contact, actions, quality time etc.).

Same here. Not thinking in words at all, very strong preference for touch or very simple expressions. Over the years with my SO, we basically formed a language of taps, hugs, noises, licks, sniffs, … (E. g. shlip tongue noise—Greetings! / I like you. / … (there are even tonal variations—rising / higher pitch is questioning, flat is affirmative), sniff-sniff noise—What’s wrong? Etc. - So a possible exchange could be seeing them sitting on the sofa, looking unhappy—sniff-sniff (What’s wrong?) - grumble (Bad mood.) - shlíp? (Affection?) - shlīp. (Yes.) - hug. That’s much less exhausting than doing the same in words.)

__nobody Dec 20, 2020, 6:10 PM
13 points
on: Gauging the conscious experience of LessWrong
Some more details on each of the categories in order:

Visual—I don’t really see things, I just get some weird topological-ish representation. E.g. if I try to imagine a cube, it’s more like the grid of a cube / wireframe instead of a real object, and it’s really stretchy-bendy and can sometimes wobble around or deform on its own. And attributes like red / a letter printed on a side etc. are not necessarily part of the face but often just floating “labels” connected by a (different kind of) line that goes “sideways” out of 3-space? o.O Even real objects like a tree are abstracted (and it depends on what I focus on), e. g. it’s more like a cylinder with a “fuzzy sphere” on top and a “fuzzy cone” below. Or maybe I focus on the bark, then that gets more detail but everything else becomes even less defined. Or maybe it’s just general “tree-ness”, then there’s just some elongated blob and I can’t even query whether it has sharp corners/edges or not without it changing shape to become more defined (and then it could end up either way, so no clue).)

Sound—sometimes, I get runaway “earworms” (“stuck tune” doesn’t really fit) that morph over time and evolve away from the source material. Sometimes there’s two of them at the same time, fighting each other or working in harmony. (If that’s (the illusion of) two full orchestras, it can get quite exhausting...) I lack the music education / knowledge to write down or play what I hear.

Taste—negligible, mostly on the “floating tags” level. “Here be salty.”

Smell—was on the same level, but getting a lot better now that I consciously care about smell in real life.

Touch—mostly negligible but sometimes I can cook up scary real stuff. (Like imagining a friend biting my neck and hairs raising up and also noticing how the muscles around that spot would move if that were real, even if I didn’t consciously follow that before, and then later validating the memory of that against the real thing. That was so surprising that I still remember it.) I find it a lot harder to imagine touch on extremities than the torso. E.g. I can’t really imagine a stubbed toe, the feeling of walking across carpet, grabbing a cabbage with my hand, but I can vividly imagine a drop of water running down my chest or a spider crawling across it. (Lower resolution is easier to fake?)

Internal monologue / thinking in words—I can’t imagine how that would be. Even when talking, I don’t really know (on the word level) what I’m going to say before I actually say it. I generally don’t think in words and have a hard time translating back and forth from words into my internal representation and back out. There’s lots of concepts that I use that I can’t really describe because they’re several abstraction levels up from stuff that maps to known external words, so I’d have to explain/translate several layers of wordless stuff before I could even describe the perspective from which these concepts arise. (And the “floating tags” mentioned above also aren’t words but just “nodes” that somehow represent the stuff.)

Memory—no clue how to assess that, I don’t have a baseline to compare against. Some stuff gets regularly forgotten (names, anyone? ^.), other stuff (like small random details) just sticks around far longer than justified. (I also tend to notice lots of weird details.)

Another potentially notable thing about memory: I can’t randomly access or scan memories, but if I have a specific query I tend to get results, including context (e.g. why this memory is probably trustworthy, stuff that helps locate it in space/time etc. etc.)

Mind control—I can’t stop an ongoing battle of two orchestras or other things that make themselves hard to ignore, but for stuff that’s not as overpowering, I can mostly just ignore it and it’s gone?

Synesthesia—I think sometimes I get whiffs of it, like 3 smelling slightly cake-y and 7 rather bloody / putting a metallic taste in the back of my mouth, but it’s rare / hard to notice. (Might be imagined, but the associations are consistent.) I also get some angle-dependent coloration on narrow grids, not sure what that is. (I know that you can supposedly trigger that effect by looking at training images and it’ll persist for a long time, but I didn’t (knowingly) look at any similar images, and also it’s doing a full 360° / full color cycle for me vs. that effect only getting you like 2 or 3 colors IIRC?)

Some other stuff:

I can feel even pretty small heat sources across fairly large distances, I notice the exaggerated impression / it presents as a tingling sensation on my face, or my hands if I’m searching with them (in which case I also notice it along my arms unless they’re covered). E.g. I can locate a small desk lamp that’s 8°C above room temperature from the other side of the room with my eyes closed and after the lamp was turned off (so I don’t react to any stray light), or notice on what side of a sofa a human has been sitting on.

Several years ago I did an experiment where I tried to do as much as possible with my eyes closed for the better part of a week. That seriously improved my ability to navigate with eyes closed. I unlearned a lot of that, but I still like to e. g. shower with eyes closed because it reduces the amount of sensory data coming in. Sometimes, when I’m stressed / close to sensory overload, closing my eyes doesn’t help and the falling drops of water create a fairly stable “image” of the bathtub / wall / curtain. (It’s like every drop’s impact creates a small gray circle that fades over ~0.2-0.5 seconds and that’s approximately aligned with the normal of the surface. Together, from all the drops, I get something resembling a picture of the surroundings. The dots are relative to my head, not the external world, so if I move my head, they move along and I get garbage until the old dots fade and things stabilize again. Aaaaaaaaaa exhausting.)

I generally have trouble navigating unknown areas, but if I reserve a few brain cycles to manually track “north” (or any direction really), that persists and I can navigate relative to that. (The marker that I use is like an “arrowless arrow”, a “shapeless shape”? It’s really weird trying to describe it. It’s pointy and has a clear designated “pointy end” but it’s not really “physical” (even in mind space). It generally floats above my head and can turn really quickly—I can jump around, whirl in place, … and it just stays glued to the target.) However, as soon as I forget to keep it active (even just for a moment) it comes loose and no longer updates relative to my movements. “Keeping it active” isn’t “thinking at it” but just observing it do its thing, don’t really know how to describe that. It also doesn’t work when I’m being moved (cars, trains, …), it seems like my guts (body location-wise) do the tracking and they don’t get (all) slow rotations.

When I do mathematical proofs, I don’t really think, I just “feel” my way through the tree of possible derivations, pick a step, see where I end up, repeat. If that doesn’t work, I get stuck. I lack the experience with manually/consciously enumerating the top layers (or this combinatorial stuff is just really exhausting for me?), and so I can basically just wait a few days and try again until I get some progress.

The same “either it works intuitively or not at all” applies to a lot of things that I do, from other technical stuff like coding to things like cooking. Usually whatever I cook tastes good to great, but sometimes it comes out weird and then I have no clue how to fix it. (That’s getting better, I’m slowly accumulating things to check / try—like aim for 0.5-1% salt (saliva is around 0.4% and anything below tends to taste weird), check if it needs more acid, add a bit of sugar, etc., but there’s still a big difference between things just working and something being not quite right.)

__nobody Dec 9, 2020, 1:54 PM
1 point
in reply to: Zvi’s comment on: Why are delicious biscuits obscure?
I suspect the water content of honey/treacle (estimating 15-20%) will lead to more gluten formation, which risks causing a chewy instead of crumbly texture. (If you’re not adding any water at all, you’re not getting gluten strands.) Butter also has some water (around 15%), so you generally don’t knead these kinds of dough for long. (Same goes for shortbread, scones, …)

Hence, I guess any flour should do if you know how to handle it / are careful not to overwork the dough.

__nobody Nov 30, 2020, 2:49 AM
4 points
in reply to: Brendan Heisler’s comment on: Final Version Perfected: An Underused Execution Algorithm
Apropos SRS and needing a second list: I’m currently experimenting with using Anki as a to-do tool. (Every task becomes a card, I “learn” the deck, and for each task get the choice of Again/Hard/Good/Easy with listed intervals, e.g. <15min, 1d, 6d, 27d. If I want to see the task again quickly (Again/Hard), I have to spend at least 5 minutes on it before clicking the button. Otherwise, it’s ok to bury it by clicking ‘Easy’ on it.)

This deals nicely with “nice-to-haves” that I don’t want to kill yet. Whenever they come up, I just send them away to next month/quarter/year/… - that way, I don’t ever have to make the hard decision to kill the idea, I could always change my mind about this later, because I know that I will be reminded eventually. These “non-tasks” are also eating a fraction of the mental capacity that they’d otherwise require, because (a) I don’t get a growing list of “things that would be nice to do eventually” that I’d have to manually comb through (instead every couple of days there will be 1 or 2 of them mixed in with the review), and (b) I know that they’ll pop back up eventually, so I don’t have to worry about keeping track of them or finding them again if I ever change my mind. (To a lesser degree, it also forces me to start early on tasks due in several months, to keep the interval growth in check.)

The only problem I had so far was tasks that are inactionable right now (e.g. taking out the trash at 5am, because noisy) needed to be ‘skipped’ and reviewed later, leading to “ugh not in the mood right now” skips of actually actionable tasks, leading to less stuff getting done.

FVP looks like it might fix this. So for the month of December I’ll be running the following modified algorithm:
1. Go through the list of tasks in Anki, for each either do it right away or put it on the (paper) list, or delegate it to future-me (i.e. pick the long interval Good/Easy option).
2. Do FVP on the paper list.
3. 3-strikes rule for “list leftovers”: If I say I do it but don’t get around to it, the counter increases. Once it hits three, I must “delegate to future me”. (Counter resets whenever I explicitly declare not to work on it, not sure yet if it should also reduce for streaks… For the experiment, I’ll go with yes and 3-days streaks for removing one mark.)
Let’s see how long I’ll stick to this.

__nobody Jul 13, 2020, 11:12 PM
10 points
on: The New Frontpage Design & Opening Tag Creation!
Some (fairly low-priority) glitches / feature requests surrounding tag filtering, in decreasing order of perceived importance.

Front page filtering / tag order: Request: Currently they’re in order of being added, with no way to reorder. I’m already getting confused by that… Manual reordering might be nice, but just some imposed order (most likely name or weight adjustment) would be better than the current state.

Front page filtering / adjusting tag weights in the popup: Glitch: When changing the tag weights via the up/down arrows next to the ‘other’ field, the size of element(s) to the left changes when reaching/leaving one of the pre-defined values, which changes where your clicks go. (Worst case, Firefox re-layouted such that the second click went to the ‘remove tag’ button… oops.) Ideally, the buttons stay where they are so I don’t have to (remember to) wait nor reach for the keyboard and manually compute the adjustment in my head, and instead can just click a few times.

All tags page: Request: Mark newly added / edited tags in some way. (I went through the whole list to add tags that I want to track. New tags are still being created all the time. Currently, it seems the only way to find all those new/updated tags is to go through the whole list again.)

__nobody Jul 5, 2020, 10:11 PM
3 points
in reply to: DanielFilan’s comment on: Site Redesign Feedback Requested
I’d actually prefer to have a true dark mode, which probably won’t be coming anytime soon… (unless the team steals the dark/light mode stuff from Gwern maybe?)

Luckily, user style sheets are a thing! There’s a dark mode style sheet from 2018 on userstyles.org and thanks to its extreme simplicity, I’m happy to report that it still works on the new site design. (Unfortunately, that simplicity also means that it’s not very good… c’est la vie. Maybe I’ll make a better one when some urgent deadline approaches some day in the following months, maybe not.)

If you prefer white, setting this as your user style sheet might get you 80%-95% of the way there:
```
.PostsItem2-background {
    background: #f4f4f4;
}
body, .Layout-main {
    background: white;
}
.PostsItem2-bottomBorder {
    border-bottom: solid 2px white;
}
```
(You may want to fiddle some more with the post / comment borders, other than that I noticed no problems.)

Of course, I’m not sure how stable those names are… So here’s some actual feedback / questions on that side of the design:
- Some CSS classes do seemingly unnecessary things (e.g..PostsItem2-isRead overrides the background color of the normal .PostsItem2-background but applies the exact same value, same for .Layout-main / body), which makes it potentially harder to customize. (Because !important in the user style overrides both classes it actually still works fine for PostsItem2-isRead, but .Layout-main needs an explicit extra rule and just deleting that background color override in the original CSS doesn’t seem to break anything.) Do you want a list of weird spots like that? (If yes, I’ll make a crazy style in 1-2 weeks and see what breaks when modding / what can (probably) safely go away.)
- stable names for most things would be really useful, but PostsItem2 looks fairly unstable / generated? How stable will those names be?
Other than that, I agree with the “looks good”!

__nobody May 29, 2020, 12:19 AM
1 point
on: __nobody’s Shortform
Observation: It should generally be safe to forbid non-termination when searching for programs/algorithms.

In practice, all useful algorithms terminate: If you know that you’re dealing with a semi-decidable thing and doing serious work, you’ll either (a) add a hard cutoff, or (b) structure the algorithm into a bounded step function and a controller that decides whether or not to run for another step. That transformation is not adding significant overhead size-wise, so you’re bound to find a terminating algorithm “near” a non-terminating one!

Sure, that slightly changes the interface – it’s now allowed to abort with “don’t know”, but that’s a transformation that you likely would have applied anyway. Even if you consider that a drawback, not having to deal with potentially non-terminating programs / being able to use a description format that cannot represent non-terminating forms should more than make up for that.

(I just noticed this while thinking about how to best write something in Coq (and deciding on termination by “fuel limit”), after AABoyles’ shortform on logical causal isolation with its tragically simple bit-flip search had recently made me think about program enumeration again…)

__nobody’s Shortform

__nobodyMay 29, 2020, 12:19 AM

2 points

1 comment LW link

__nobody May 26, 2020, 2:52 AM
34 points
on: Baking is Not a Ritual
The key point is this: The big difference between baking and cooking is that baking involves much more chemistry than cooking, which means that altering the recipe without understanding what you’re doing is much more likely to result in failure. (Bad substitutions/ratios in cooking means the result might taste/look a bit strange, but overall it’ll likely be fairly close to the original. Bad substitutions/ratios in baking means you probably get a brick / dust / other unidentifiable garbage.)

Thus, if people approach baking like cooking, they probably fail. Repeatedly. Hence the ritual thinking.

__nobody May 24, 2020, 4:19 PM
4 points
on: What are objects that have made your life better?
Erasable gel ink pens in lots of different colors.

Working on paper still beats tablets etc. sometimes, and instead of crossing out stuff and trying again, you erase and overwrite – looks much cleaner, even if this was your very first draft / rough diagram / whatever. Instead of copying / re-writing the whole thing a half-dozen times or more to get a final clean version, you copy maybe once, at most twice, often not at all. (You just erase small mistakes that happen while making the clean version, instead of starting over yet again.)

Beats colored pencils by a wide margin, both in handling when writing/drawing as well as in ease of erasure. (The ink becomes completely invisible when heated, no need to scrape pigments out of the crevices of the paper / abrade the paper surface.)

Muji (the “Japanese Ikea”) had great ones, but they got rid of most colors (no more green/cyan/purple/…, only black/red/blue.) Luckily, lots of others are producing them now, so I can get new ones when mine (and their refills) finally run out.

My only warning: If you’re writing double-sided in a notebook with thin paper, don’t be too vigorous when erasing. Normal corrections are no problem, but taking out a whole shaded diagram might also erase parts on the back. Other than that, while I’m not sure how long-term stable these inks are, my 5+ year old notes still look fresh. (I still made backup photos just in case…)

__nobody May 8, 2020, 2:49 AM
5 points
on: Value of building an online “knowledge web”
I haven’t created an account on their page, so this is based purely on what I’m seeing in the example collection / demo videos. It looks broadly similar enough to what I’ve been building/using over the last years[1] that I think a summary of my experiences with my own tool and the features that you will need might be useful:

In short: It looks awesome as long as you have only small amounts of content – and I think it may actually be awesome in those cases: For almost all my projects, I’m creating separate(!) maps/webs and collecting todos and their states to get a visual overview, and these are also useful to get back into the project if I come back months or years later – so they’d probably also help other people too. But… as things grow, it will get very confusing. (Unless you add lots of functionality into the tool and put in extra effort specifically to counter that.) All attempts to collect all my projects / ideas in a single big map have failed so far… That said, I haven’t given up yet and am still trying to make it work.

Here’s how things have been breaking down for me:

naming things is hard and over time, you’ll pick subtly different names (unless you have a fast way to look up what you called the thing a couple of months ago), and then spelling variants will point to different pages and the web breaks down… Also, there will be drift / shift in meaning or goals – you explore a new sub-topic and suddenly it looks like a good idea to rename / move a page or even a whole category, in sync, across all pages. Without tool support for both of these, things will fail. (p=1.0, N=3; after adding fast search (not just prefix-based like autocomplete!), the 4th map/web died of ontology instead of spelling variants and the 5th isn’t dead yet…) Name changes are made simpler if you have page IDs that are independent from the name (and actually Roam has that, too – so these two shouldn’t be a problem. A difference is that they’re using auto-generated IDs whereas I’m manually choosing mine and using them to encode hierarchy or other stuff that I want in there… I personally find that useful (as long as you don’t over-ontologize), but YMMV….)

Next is unweighted linking of everything: If you see all the gazillion uses, sub-topics, … of linear algebra when you look at that page, that’s not much better than seeing none of them. (Same for visual graphs.) Automatic sorting (or even just highlighting) by centrality / importance or, failing that, manually curated sub-lists are a necessity. Related, there’s not hiding stuff outside the local project/topic: Separate small maps / collections, that can still refer to each other but are largely independent, seem to work better than having everything in one big blob. (Technically, that’s basically equivalent to assigning a unique name/ID prefix to each project, but having to manually add that everywhere is draining. It’s bad enough that I started working on splitting things up, in spite of all the new problems that brings up… To name just one, things aren’t truly separate at the mechanical level, as you’ll still have to rename/refactor in sync.) Also, graphs can pick a better layout if they’re not constrained by the placement of nodes that you don’t want to see anyway.

Beyond that point, I don’t know yet… the last two are still only partially implemented. (And I’m not seeing anything like these in Roam… so you’ll probably run into problems there eventually.) So far, it looks like that might be enough and from that point on experience in steering / organizing the thing becomes more important.[2]

On graphs: What I’m seeing in Roam is… underwhelming. (Same for the example linked by mr-hire.) Unless I’ve looked at bad examples, the graphs that Roam gets you are a fixed grid-based unordered (i.e. not ordered to minimize total edge lengths or something like that) mess and the only thing you have is that you can click a node and see its immediate neighbors highlighted? (Or the per-page graphs are essentially a linear list, presented as a very wide 1-deep “graph”?)

If that’s what you’re working with, then of course that won’t be an effective learning tool. When I read you talking about the “clarity” of seeing the connectedness I thought what you’d get from Roam is much closer to something like this:

This is a small-ish part of an older map of mine (with node labels censored)… rectangular nodes and the wide blue arrows show hierarchy, dotted arrows (not sure if there are any…) are references/mentions, solid arrows are dependencies. Green stuff is done, light beige stuff is dead/failed. The bright orange bubbles are temporary(!) todo markers – what do I want to achieve / understand now? The fat / colored bubbles are the response (a gradient from bright red thru purple and blue fading to white) – this is what you can work on next and how important it is. (Thin / gray nodes are off the active paths and can probably be ignored, nodes on the paths with open dependencies also show the max. flow through them in their border color so you can decide to cut further dependencies and bodge something in place of that sub-tree.) As you can mark multiple things at the same time (also with different weights/priorities), you can not only get local “to get X, you’ll need A, B, C, …” information (plus information on the branching and where it might make sense to cut…) but you can also get “broad field” information. (“X, Y, Z all indirectly/weakly depend on A”, so working on A might simplify work on all of these – something that you might not notice if you look at X, then Y, then Z one after another…)

If you have something like that – where you can ask questions and get responses based on topical dependencies and what you already know, and iterate those questions – I think that can be an effective learning tool. But I don’t know for sure yet, this thing doesn’t work well enough yet to manage large amounts of nodes/information… (My gut feeling currently says that you want to split things very fine-grained – every concept / theorem is its own node, so that you can mark them as done independently – and then when you work through, say, linear algebra, that’ll be a lot of nodes. Not there yet, still not enough hiding/filtering…)

(Another thing I’m undecided on is whether I want/need “xor nodes” – “to get X, you’ll need A or B or C; not A and B and C” – that might allow much fancier optimization but it also takes agency/information away from you and I’m not sure that you’d get the best decisions out of that, especially if map information is partial/wrong/incomplete and all that…)

As for why this isn’t a thing yet… I guess it’s (a) hard to make something that actually works well and lots of people try a bit and give up and then others see all those bad examples and conclude it can’t work, so less people actually really try? Also, if you finally had the tooling, (b) getting it right would involve a lot of data entry, and you’d need more experience than you’d need to edit Wikipedia (I suspect closer to something like Wikidata) but it’d be a lot less work to just write a linear text book and be done with it. And (c) machine learning can help with the tagging involved in (b) and simplify that down to Wikipedia-level, but the fancy stuff has only happened in the last couple of years (and it’d be even more work on the tooling side)… so it’s possible that something exists somewhere that works fairly well, but it’s still rather unlikely?

[1] It’s a terrible terrible terrible “organically grown” ~1.5KLoC Lua script (about 20% of that in a single function…) that, via Graphviz and Pandoc, generates (static) colorful graphs and HTML pages. Primary focus is tracking / prioritizing todos and projects, but that includes learning new stuff and recording knowledge. (Especially when approaching new areas of math, you tend to get loops (to understand A, you want to understand B, which can be done via C, which relies on A…) and one of the tasks will be to break up those loops… None of the existing tools I found were able to record and work with that. So that’s how this started…)

[2] E.g. for my planning thing I now have broad categories like “knowledge” (static, non-actionable, non-directed), “projects” (active, non-actionable, directed / actively carving actionable todos), “farts” (inactive projects, ideas to do stuff, …), “beacons” (fuzzy long-term goals / directions to move in, grouping many of the projects), “spikes” (actionable project carvings) and specific (sub-)projects/tasks are repeatedly moving between these – ‘projects.foo’ grows a spike somewhere in its text, it moves to ‘spikes.foo.how_random_is_enough’, if I stop working on it it moves back into the project (and once restarted back to spikes…), and when done it gets a write-up and moves to ‘knowledge.foo.random.spike’ (for archival purposes), plus extra nodes like ‘knowledge.foo.random.lcg_is_not_enough’, ‘knowledge.foo.random.pcg_works’, … (for fast knowledge access). So far, this seems to finally reach the point where it starts to work… for me.