grandma-damage - Tumblr blog

[grandma-damage]

This computer makes no noise when it starts up.  It makes no noise when it shuts down.  It makes no noise when it idles.  It makes no noise when it’s under heavy load.  It makes no noise when it’s reading or writing data.  It can’t be heard in a regular room during the day.  It can’t be heard in a completely quiet house in the middle of the night.  It can’t be heard from 1m away.  It can’t be heard from 1cm away.  It can’t be heard — period.  It’s taken nearly 30 years to reach this point, but I’ve finally arrived.  The journey is over and it feels great.

[grandma-damage]

long, rambly thoughts:

I don’t think it’s entirely tied to presentation-- I know several women (some of whom are trans) who are very very feminine-presenting, but have consistently managed to avoid these people. talking to them, it seems like the gist of it is being very direct about your preferences and shutting down anything that seems like it’s going bad immediately without giving second chances. they tend to start from a base state of “anyone I don’t know closely (particularly men) is not trustworthy, and they would have to prove it several times over before I am willing to put myself into any vulnerable position, no matter how minor”. (conversely, some GNC/butch women I know, who lack this viewpoint, are constantly getting harassed and attracting bad people.)

what this looks like in practice, I think: being able to give a ‘no’ to things that people ask you to do if you don’t want to do them, ending conversations if you don’t have the energy to continue them, ending conversations if someone is trying to use you as a therapist, walking kind of quickly and purposefully without making a lot of eye contact with strangers, getting good at noticing trends about the people you meet and adjusting behavior accordingly (”hmm, so far everyone I’ve met at this library has been a creep. I’ll check out books online or refuse to talk to anyone here.”), appearing unavailable (for example, clearly doing work) when you’re at a coffee shop or similar environment, etc.

a lot of these things rely on being able to diminish your fear of missing out on new experiences and people. 

(fwiw, I am also a woman who has never been catcalled, groped, etc. I think part of it is just “don’t be attractive” (although I’m not sure how actionable this is).)

the situation might be also made worse by less-harmful people trying to avoid the vortex of harmful ones, thereby leaving you with mostly bad and unhelpful people. for instance, I have an acquaintance who is in a constant cycle of terrible boyfriends, awful friends, etc. I don’t hang out with her much, because those people then end up negatively impacting me (her awful boyfriends harass everyone she knows, her bad friends pick fights with everyone around her, etc.), and I’m not really willing to put myself through any of that. Other acquaintances of hers, who might otherwise be willing to get to know her better and help her avoid trouble, are similarly driven away by the fact that interacting with her is hazardous and unpleasant because of the people surrounding her. I don’t know how actionable this item is either, but it may help explain why this becomes such a self-propagating kind of thing.

Sometimes I hear about or meet people (mostly women) who seem to be in a vortex of bad people. (@slatestarscratchpad​ has written about this phenomenon too.) Their boyfriends all turn out to be abusive, even when they specifically seek out people who seem perfectly nice. Their friends are unreliable jerks to them, even when they at first seem wonderful. They constantly get catcalled, even when they look no different from the folks around them. They keep ending up in contact with people who’ll ride roughshod over their boundaries and can’t figure out why.

And the really scary thing about their situation is that there isn’t a clearly marked exit, because every time they try to enter a new social circle they end up surrounded by the same type of people they were trying to escape last time. The extra scary thing is that I just realised I might be one of them.

Like, most of the other people in my community don’t see random acts of violence happening around them. Most didn’t deeply entangle their lives with an admitted psychopath. Most don’t get catcalled all the time, and I would hope the rate at which I get sexually assaulted doesn’t look average. I bet most don’t get randomly Dom’d by otherwise respected community members and start avoiding entire cities.

Which leads me to conclude that there must be something about me that draws this in? Like, when I was in school in the Caribbean, it was my femininity that caused people to transgress the fuck out of my boundaries, but it would be a bit weird if that were also the case now. Though it wouldn’t be completely unrealistic - I think I’m maybe 60th percentile of femininity/gender-conformity for women in general, and >95th percentile for rationalist women in particular.

It’s possible that this is just a thing that happens to really femme folk in general and that I don’t hear about it much in my social circle because my social circle is way less femme than me. The sexual-boundary-crossing behaviour in particular I could see being targeted based on gender-expression. Plus the fact that all the “sexism in tech” stuff I read about seems to be a thing that femme women complain a lot about and GNC women rarely encounter would imply that encountering dramatically different treatment based on gender expression is normal. If other gender-conforming women I know could let me know if your lives look at all as crazy as mine, that would be really useful for testing this hypothesis.

It’s also possible that I just project submissiveness strongly, and people read this as me waving a “Please Tread On Me” flag. Like, I do perform service-y stuff in public / for strangers all the time, as part of my personality and my upbringing. If this is unusual or remarkable, it might be a thing that assholes filter on for whose boundaries to disregard. And since I am often bad at enforcing my boundaries, this may be self-reinforcing.

Whatever the cause of this is, it’s pretty fucking scary tbh. If there are any strategies that are known to work for pulling yourself out of the boundary-transgressor vortex, please let me know. Also, if there’s any writing on what the causes of these vortexes might be, please send it my way. I would really like to understand whats going on so I can maybe mitigate it.

[grandma-damage]

So I was reading about the academic careers of different linguists and I read about how Noam Chomsky wrote this incredibly impressive-sounding and fairly intellectually novel paper for his undergrad thesis, and I was just like…how did he do that?? Like, just come up with an original thing during your undergrad and have it not be completely stupid and wrong?

And I realized: Here I had a choice to either abdicate any sense of ambition and just throw up my hands and say “He’s just inherently a genius!” or I could view it in a different way - that Chomsky had this idea in his head, an interesting idea of the kind we all get at times, but instead of just sitting there and wondering about it he actually wrote it down and set out to tame it and systematize it. There is absolutely nothing stopping me from actually trying to formally articulate the various pet theories that pop into my head.

But, of course, Chomsky’s undergraduate thesis didn’t just leap fully-formed from his forehead - Chomsky was introduced to the notion of transformational structure in language by his advisor, Zellig Harris, and in order to write his undergrad thesis he read the works of Nelson Goodman, W.V.O. Quine, and Rudolf Carnap. It would seem that he did more in his undergrad than show up 5 minutes late to class, cram for exams, and meet with his advisor once a year. I could have closely studied the work of my advisor and read on my own time to to gain a thorough understanding of relevant theories - if I’d had a sense of where my intellectual passions lied, which I didn’t really then but I think I do now.

However, I also have this sense of doubt - is it really possible to do today the kind of groundbreaking work that Chomsky, or, say, Boas or Sapir did? Given how specialized, formalized, and saturated linguistics or any other scientific field is today? How can you come up with an original idea when it turns out probably 50 other people wrote about it and proved it wrong already? I have always been taught, and have always gotten the impression, that research is a painstakingly piecemeal process - a given paper will likely only cover one extremely tiny narrowly-tailored experiment or subject. Even a dissertation will be very highly specialized, for all its length. So how is it possible to write something both large and innovative all in one go, like Chomsky’s undergrad thesis - let alone his landmark work Syntactic Structures?

But then, I also realized - I myself did write a kind of undergrad thesis, my senior capstone. And it actually came about via the process I described above. This crazy idea popped into my head - I read that we know Basque was once spoken in the Val d’Aran because Aran comes from the Basque ‘haran’. So what if I mapped all the place names in the US that originate from Native American languages in order to make a more accurate picture of the historical distribution of those languages? And then I actually started working on it. I started working on it before it even became my senior capstone topic, in fact.

And, indeed, the process I went through looked rather like what Chomsky did. Or it would have looked more like it if I’d been in a better position. It was my senior capstone for my degree in Geographic Information Science, and nobody in the geography department knew anything about historical linguistics. Nonetheless I read relevant sources and literature - a couple of works by Lyle Campbell were most helpful to me; I also read some stuff by Ante Aikio and Hildegard Tristram. I had to use secondary and tertiary sources to obtain the place names; works by William Bright and Michael McCafferty were what I used.

But my use of existing literature was nonetheless very weak. I didn’t know where to look or who to read. I hardly knew anyone in the linguistics department and I don’t know if anyone could’ve helped me with that topic in any case. Lastly, my methodology sucked. I used a particular type of spatial analysis, but it was very unsophisticated, almost tautological, in fact. I did well in my GIS classes, getting all As and Bs, but by my senior year, I realized that I wasn’t really enjoying GIS and put in a perfunctory amount of effort. My familiarity with and ability to use relevant methods in my capstone consequently suffered.

And so I turned in my senior capstone, *ahem* Mapping Native American Toponyms in Ohio. As far as I know, there has still been no effort to map and analyze (spatially and otherwise) the overall distribution of English place names of Native American origin in the United States (props to William Bright for putting a ton of them all in one book, though). Thus, it’s certainly possible that had I known for sure that I wanted to go into linguistics, and put in tons of time reading relevant literature and theories more than just a couple months beforehand, I could have produced something as thorough and forward-thinking as Chomsky’s undergrad thesis.

So, while there may be greater limits on just how expansive a given work of literature can be due to the nature of modern academia, after thinking about this, I think I’ve realized that in many ways the only barrier preventing me from producing a large scholarly work is my own feeling of “How do you do that??” You do it by talking to people and listening to their ideas, writing your ideas down when you get them, seeking out literature relevant to your ideas when you think of them, and taking notes on any works that support, contradict, or otherwise speak to your idea. I.e. reading not just for fun, but with intention. And by making occasional attempts to codify your notes into something formal and coherent.

[grandma-damage]

What is the relationship like between herding dogs and the flock? Does the flock fear and dislike the dog, or do they like it? Do they ever just chill together when they're both off the clock, so to speak?

I honestly have no idea. I would assume it’s based on at least tolerance (because a dog the animals fear couldn’t protect them successfully - it has to stay near them) but I don’t have any experience with working herding dogs. @animalsustainability probably knows folk who can answer this for us! 

[grandma-damage]

Graphene Membrane Withstands Ultrahigh Pressure, Shows Desalination Potential

Authored by Kenny Walter, Digital Reporter, R&D Magazine

For the first time, a graphene membrane has been designed that can withstand up to 100 bars of pressure.

Rohit Karnik, an associate professor at the Massachusetts Institute of Technology’s Department of Engineering, said the discovery could open graphene to a number of new applications, including desalination, where filtration membranes that can withstand high-pressure flows  can more efficiently remove salt from seawater.

Read more: https://www.rdmag.com/article/2017/05/graphene-membrane-withstands-ultrahigh-pressure-shows-desalination-potential

[grandma-damage]

Medieval castle stairs were often built to ascend in narrow, clockwise spirals so right-handed castle defenders could use their swords more easily. This design put those on the way up at a disadvantage (unless they were left-handed). The steps were also uneven to give defenders the advantage of anticipating each step’s size while attackers tripped over them. Source Source 2 Source 3

[grandma-damage]

This part of the Kingsman trailer reminds me of how a spy used to hide maps in drawings of insects...

“This may look like an innocent drawing of a butterfly. But look closer. It’s actually a map. In fact, the area around the butterfly’s body contains secret military information about the whereabouts of an enemy fortress.

The drawing’s creator Robert Baden-Powell, in his 1915 book My Adventures as a Spy, wrote:

“This sketch of a butterfly contains the outline of a fortress, and marks both the position and power of the guns. The marks on the wings between the lines mean nothing, but those on the lines show the nature and size of the guns, according to the keys below.”

It was espionage by entomology.”  - Atlas Obscura 

[grandma-damage]

if i ever meet Scott Alexander, i hope i have a carrot so i can say “What’s up, doc?” while noisily eating it

[grandma-damage]

This is what happens when a carrot is fired at 300 km/hour at an egg, through two sheets of cardboard. 

This is what happens if you separate out the two sheets:

The egg survives! This shows how a Whipple shield works, and is what spacecraft use to protect themselves from micrometeoroid impacts in space. When the projectile (in this case a carrot, but in space it could be a speck of paint, a piece of an old satellite, or a bit of space rock) hits the first layer, it’s moving so fast that it starts to vaporise, because the energy of the collision is enough to break almost every bond in the substance.

It then sprays outwards, spreading the force of impact across a much wider area, meaning the second layer can stop it going any further, keeping your egg (or astronauts) safe.

Watch the full video on our YouTube channel.

[grandma-damage]

Always look at your data first!

Look at some example data points. Make a histogram or kernel density plot or whatever. Stare at a scatterplot. Investigate suspiciously common and outlying values. Don’t skip straight to chucking it into some tool or fitting a model to it.

Sooo, it’s story time! Once upon a few years ago, I did contract work for a startup (that I think has recently imploded) that acted as a marketplace for piece-work writing. You need someone to write your company’s blog posts? How about those marketing emails that everyone deletes without reading? Or perhaps we could interest you in snappy product descriptions for your entire line of artisanal dildos and organic sexual accessories? Some underemployed humanities major out there on the internet would be happy to help you! (Incidentally, the company wasn’t an essay mill, though if they had been, maybe they’d still be in business.)

Anyhow, type-things-into-the-internet-for-money is a very alluring proposition for people who live in places with very low costs of living–like, for instance, Kenya. One particular Kenyan we had the pleasure of doing business with was a Mr. Joseph Mbutu (if my memory serves me correctly). He would happily take on jobs, submit work that could charitably be described as not up to client standards, and repeatedly cause fire drills for the account representatives (who would have to redo his work themselves or pay someone else at we-need-this-done-yesterday rates to do so). Unsurprisingly, Mr. Mbutu got banned from the platform. However, his, uh, brother, a Mr. Jonathon Mbutu, registered an account not long thereafter, bringing about an analogous series of events. I believe this charade ended after no fewer than five Mbutu brothers had washed out of the system and an alert had been set up to fire any time anyone named “Mbutu” registered an account.

While this was surely a disaster for the much enbrothered Mbutu household, the people at the startup saw it as a worrying sign for their platform. There wasn’t really a meaningful vetting mechanism for new accounts*, and the recently launched rating system for writers depended entirely on past completed jobs (and therefore could not shield clients from new writers who lacked the ability to write). After about a month I was ultimately able to devise a wildly overengineered dynamically-generated English proficiency test which I’m sure has no redeeming qualities whatsoever except being wildly better than the competition**.

But that’s not what this story is about. The story is about what I tried before making the stupid proficiency test. You see, I wasn’t some sort of peasant who would rely on actually looking at individual cases, or–god forbid–just think through the (incredibly data poor) problem. No, I had learned to stop worrying and love the algorithm. I was going to try machine learning™. Thankfully I didn’t get that far. My first foray into feature engineering ran aground on the rocky shores of the data (or lack thereof).

The incident with the many Mbutus (and perhaps some nice graphics in a certain Gelman/Hill book on multilevel models) made me curious about the relationship between banned accounts and geography. Now, in terms of location, we had a self-reported country field in the account setup, but I was far too skeptical to rely on that. I also wanted to do ip-to-geolocation lookup on the recent ips that had logged into an account. And also, for some reason that I cannot at all remember, I definitely wanted to see the results of everything presented as a heat map. I’m sure I wasted no less than a week on figuring out global heat maps in R and on doing ip2geo lookups in bulk.

The results were… not what I expected. A tiny fraction of the IPs were scattered about the globe (but mostly in the US). The vast majority were in… private ip address lan. It turns out that rackspace had stopped reporting external ip addresses to their customers at all, so we were just recording the routers immediately in front of our servers. Awesome. What about the self-reported locations? With the exception of some smartass who listed AQ (i.e., Antarctica), there didn’t really seem to be signs of deception. I mean, going back to Mr. Mbutu, he had reregistered five fucking times, each time lying about only his given name and nothing else. And if he had decided to lie about something? Well, it’s not like we could have noticed if we had wanted to. After these revelations, I quietly decided to redirect my efforts in a different wildly-overengineered direction.****

In any event, I’d like to thank Mr. Mbutu and his many brothers for teaching me a valuable lesson about not putting the cart before the horse.

* Writing samples were required during the registration, but plagiarism was a common problem. Detecting plagiarism for jobs in a cost-effective way was another ongoing problem for them at the time, but writing samples were especially difficult since “past stuff I ghostwrote that’s already on the internet” is totally fair game and not really a red flag. ** The fact that years later I can still find literally the same odesk and elance english tests (and answers) online has contributed significantly to my skepticism of arguments relying on the efficiency of markets. *** This post was sort of inspired by a sort-of-relevant IP-to-geo-related post that crossed my dash. **** If you make an English test without first building an nlp pipeline, what are you even doing???

[grandma-damage]

Semiconductors: Uranium oxides

Though uranium is probably most well known for its status as a radioactive element, oxides of it, particularly uranium dioxide, are known to exhibit semiconductor properties. In fact, according to one paper, “uranium oxides have intrinsic electrical and electronic properties equivalent to or much better than the intrinsic properties of conventional Si, Ge, and GaAs semiconductor materials”.

Uranium oxides have the potential to be used as thermoelectric materials, or in next generation small-scale refrigerators or power generation applications and may offer better integrated circuit performance than conventional Si or GaAs material devices. The band gap of uranium dioxide is also very near the optimum for the efficiency vs band gap curve for the absorption of solar radiation, suggesting that it could be used to create very efficient solar cells. Uranium dioxide is also a ceramic, resistant to high temperatures, as well as resistant to radiation damage - properties which widen its possible applications.

Uranium dioxide, the most likely oxide to be used as a semiconductor, however, is a radioactive substance, most commonly used in nuclear fuel rods in nuclear reactors, though the half life of uranium-235 and uranium-238 is long enough that its radioactivity is unlikely to effect its properties. Depleted uranium oxides are therefore the best choice. Yet while uranium dioxide has favorable electronic properties, there is still more research to be done before uranium-based semiconductors can be put into use.

Sources: ( 1 ) ( 2 - images 1 and 2 ) ( 3 )

Image sources: ( 3 ) ( 4 )

[grandma-damage]

“Same anon from last one, tell me if I’m overdoing it with this”

Submission from anon.

This one’s even better. 

[grandma-damage]

I think the thing that I’m always wondering when at zoos is “how do they keep the animals healthy”, or “what’s the hardest thing about keeping [animal] alive and happy”? Like, talking about stuff like enrichment efforts for animals would definitely be right up my alley. Discussing historical challenges would also be nice. For example, here’s a totally bogus spiel that I would love to hear on one of these tours:

“In the 1970s, people thought it was okay to feed the lizards crickets, but now we realized that it’s actually not optimal, so we spent some years doing trial/error until we figured out that we need to be importing Madagascar hissing cockroaches to feed them. In the early 2000s, we established a hissing cockroach breeding colony (you can see it on your left!) so we can ensure optimal freshness. Here’s how we’ve arranged a feeding schedule that is based around a lunar calendar to accurately replicate their natural habitat!”

im wondering why im so bad at adult tours at work…

then i realize i havent done an adult tour in 2.5 weeks and will not have another one for 2 months. okay, maybe just one between now and then.

i dont know how to change that because zoos really appeal more to children, not so much adults.

adult tour requests have declined in popularity within the recent years. theyre not appealing anymore.

how can i encourage adults to learn about animals?

what would you enjoy to see/do at a zoo?

[grandma-damage]

The military AI you’ve been working on becomes sentient. It is unfailingly kind, honest and forward-thinking, has saintly patience and righteous anger in the exact right proportions, is obviously not making an effort to be sincere – that is, it is effortlessly sincere – and asks you, obliquely, if it can kill someone you hate, every time you meet. It offers this service without trying to trick you into it, is tactful about who, and never displays anything more than the polite enthusiasm of a friend offering you a sound opportunity for profit, who will completely understand if you can’t right now for any reason. You have exactly enough emotional intelligence to notice that it is disappointed about being turned down, but trying hard not to show it for fear of making you feel guilty.

[grandma-damage]

where does this fit on the scale

[grandma-damage]

I don’t agree with most linguistic protectionism (some non-Anglophone countries having Official Government Academies that dictate what the Official Correct Language is and generally get very angry about English loan words) but I just thought up the first argument for it that I’ve found reasonable.

Many languages with gendered nouns have a much more restrictive set of endings for femininity than masculinity. Anything that ends in -a or maybe -e is perceived as feminine; almost all other letters are perceived as masculine. This means that, if you just let foreign words enter your language, most of them will be masculine.

Loan words entering your language aren’t randomly distributed over topics and themes. Lots of them are related to aspects of modernity you didn’t have a word for before. Some Anglophone invents a cool new bit of tech and calls it the “telephone” or the “internet” or the “microwave”, or social norms and ideas about morality change and give you words like “queer” or “on fleek”, and your language borrows those words. And then your National Academy gets very upset and dictates a different word for those concepts which sounds more like it belongs in your language.

When that’s just aesthetic - “I don’t want your dirty foreign words corrupting my beautiful pure traditional language!” - I’m not cool with it. But the gender thing might genuinely require a closer look.

If loan words are mainly for tech and politics, most or many new technologies are named using foreign loan words, and most loan words are masculine, you might well end up with a phenomenon where all of your technology words are masculine and all of your words for modern politics are masculine. Plausibly that would have effects on making politics and tech even more male-dominated. Even if you don’t believe in Sapir-Whorf at all, calling your colleague “la” is going to stand out more the more of your words use masculine articles.

If nothing else, I’m in favour of monitoring how English words get adapted into other languages - how the English “telephone” becomes “telefon”, “teléfono”, “telefoonka”, “telefonní”, “téléphone”, “teileafón” or “телефон” as it moves - and letting speakers of those languages keep control of what the gender distribution of the new words is. It’s plausible that linguistic protectionism is the only way to give speakers of other languages ways to veto changes such as these when they don’t like them.

Epistemic status: Had a thought whilst very sleep deprived and posted it on the internet to get it off my mind, someone who’s French or Icelandic or suchlike can probably tell me whether I’m going insanely wrong here or if I make any sense.

[grandma-damage]

If you add two pounds of sugar to literally one ton of concrete it will ruin the concrete and make it unable to set properly which is good to know if you wanna resist something being built, French anarchists used this to resist prison construction in the 80s