Obviously not what she was saying but still glad to see we're back on that "What if we just shoot the hurricane?" shit

I mean

if we have the technology to build a hurricane cannon

why not?

The Six Million Dollar Man Fan

Good 'ol Java

Oh good we've angered the cosmos as welll

@TimStone what does this even mean?

do we fly when it gets to a G6?

@GodEmperorDune swpc.noaa.gov/noaa-scales-explanation

> some protective systems will mistakenly trip out

Boo dsa swag shop is sold out

Closed until October

Hahahahahahahahahahahaha

wait what did he threaten clinton with?

I dunno

A bizarre man made a bizarre threat probably

Meanwhile

Elizabeth Warren backs Medicare For All

I would legitimately vote for Elizabeth Warren 2020 and get reasonably excited about that as a thing

But that's Kamala Harris and Elizabeth Warren on our side that we know of so that's pretty cool

Ooo, my boy Sherrod Brown is not decided

He's up in 18 so I honestly wonder

Ok Google remind me to call Sherrod Brown tomorrow

@Unionhawk ...

mcsweeneys.net/articles/…

My headcanon is he hated every minute of dunkirk

the reality is that the guardian just picked an image of him happy and he didn't for some reason but shhhhhhhhhhhhh

spontaneously combusts

@Unionhawk is there ever a time that pruitt wants to talk climate change?

@fredley hue

@GodEmperorDune you can only talk about it on "Climate Change Is Not Real Tuesdays"

which is for some reason a federal holiday

"election integrity" commission lead incorrectly accuses thousands of people in New Hampshire of voting illegally in 2016

love too watch voter supression happen

(This is also why you shouldn't be writing breitbart opinion pieces when you are working in govt)

Basically they said look at all these people who voted in New Hampshire but don't have state driving licenses

That doesn't mean they voted illegally by any stretch

That's not even a requirement in many states you just need to live there you gajhfkljsdhfg'

fuck

@Unionhawk It's a good thing I don't have RO here because I would have pinned that weeks ago

clearly the solution is to only allow people with nonexpired passports to vote

Seriously

It is a "let's disenfranchise college students from out of state" ploy

YOU SEE

PEOPLE FROM OUT OF STATE ARE COMING HERE

AND USING THE THINLY VEILED EXCUSE OF "LIVING IN THE STATE IN NOVEMBER" TO REGISTER TO VOTE HERE

IT'S ALL A COLLEGE DEMOCRAT PLOT

I MEAN

UH

ILLEGAL VOTING

BY, UH, NONCITIZENS

CLEARLY

The only people who should be allowed to vote are people who registered to vote in 1984 and who voted for Reagan

Voter registration rules are set by the states, right?

@MBraedley yes

But there are federal rules that basically say that you can only vote once in a federal election, right?

(in addition to the state rules that say the same)

you are def only allowed to vote once in an election, regardless of states, but i'm not sure if that is a federal rule or a rule that all the states have

The point I'm getting at is that in university, I was registered to vote both in my parent's district as well as the district that my university is in, which was totally legal, but Elections Canada rules said I could only vote in one of them.

Some of these New Hampshire students might be in the same situation, if the NH rules and their home state rules don't disallow being registered in multiple districts.

@MBraedley i can't speak specifically for that state, but most states are fine with you being registered in multiple places, as long as you only vote in one

part of it is that when you move, you register in New State but Old State never gets any notification to cancel it's version of your registration

journalists pointed out that this scenario affected many people high up in the trump admin, including some of the trump kids

nytimes.com/2017/01/27/us/politics/…

Excellent

hue

Honestly AVR now

@GodEmperorDune Yeah, we got mail to our house from the ODP from somebody who graduated more than 2 years ago

Like "hey we noticed you haven't voted lately"

So this is technically a potential issue

But practically speaking we don't have busloads of people going across state lines to vote more than once

Like that's not a real thing that happens

and voting more than once is 1) easily discovered and 2) severely punished

and 3) unlikely to swing an election one way or the other

There are 4 credible cases of somebody attempting to vote more than once in 2016

Except this was never about real things this was about a top-down destruction of voting rights in this country

well yes

under the guise of false voter fraud claims

> Nearly half of Republicans have yet to hold a public town hall since Donald Trump took office.

I wonder how that compares to Democrats while Obama was in office or while they were trying to pass Obamacare.

@Yuuki Yeah, I saw the same thing when I saw that headline. That number is meaningless without a basis for comparison.

I mean, anecdotally speaking, I don't remember much hubbub about representatives not holding town hall meetings.

There was a lot of talk thrown around about the bill itself but not much on the availability of town halls in particular.

There was a lot of noise about how they rushed it through in closed door meetings which is objectively not true

Especially not compared with how AHCA etc was done

mueller wants to interview spicey, the paranoiac, and hope hicks

lmao at this live election night show

"I think it's going to be trump" hours pass "Oh fuck this is actually happening"

A neural network was developed which can correctly distinguish between gay and straight men 81% of the time, and 74% for women

tldr: The research found that gay men and women tended to have “gender-atypical” features, expressions and “grooming styles”, essentially meaning gay men appeared more feminine and vice versa. The data also identified certain trends, including that gay men had narrower jaws, longer noses and larger foreheads than straight men, and that gay women had larger jaws and smaller foreheads compared to straight women.

so it probably rates all trans people as gay probably which is just 👌

Neural networks in general basically just find statistical correlations. That one seems to have found statistically significant correlations between reported sexual orientation and certain aspects of facial appearance.

Don't put more stock in it than that. It's just a classifier with a decent accuracy rate; it's not making judgments

ofc

it's more of an internal feeling you should discover than something external that can be classified from your outlook, so I'm kinda surprised at 81% tbh

@Avery there was a good explainer of why this is not that great a result yet

one moment

> If the computer can tell the difference between gay and straight people, it might tell us more about profile pic choices than actual faces

taking pictures of people in a controlled manner with same pose, same place, same expression etc might've provided more average results imo

Yeah, the article I read mentioned that the researchers acknowledged in the paper that the data was from a dating site, and that that probably affected the accuracy

Don't think that the paper follows the scientific methodology that well in that aspect.

@GodEmperorDune You think people would do that? Go on an Internet dating site and lie?

@Avery As long as they acknowledge possible confounding factors, I think that follows the scientific spirit perfectly well.

hm, yeah, you do have a point

@Yuuki I think it's not "lying" that's the issue here. More that people might be signalling their orientation to prospective partners (purposely or subconsciously)

And also sing it with me, "~correlation does not equal causation~".

@murgatroid99 yes

like if you are trying to attract a certain subset of the population, you use a profile pic catered towards that subset

if you are trying to attact a book nerd (for a trivial example) you post a pic of yourself in front of filled bookshelves

On the other hand, if the points of correlation include stuff like facial structure, then signalling isn't the whole story

@GodEmperorDune Well, the article seems to claim that they did specifically analyze facial structure features

The analysis involved ~35,000 images, so the probability of spurious correlation seems relatively low.

Hmm. I do have doubts about their methodology for determining what parts of the image were relevant to classification

For each image, they basically took a 32x32 grid, and re-ran the classifier with each square blacked out one at a time. This seems to ignore both the expectation of correlations between parts of a face, and the fact that neural network classifiers don't really work like that

Well, they also did some other analysis, based on averaging the "facial landmarks" of the most strongly classified faces on each end

@Yuuki I don't think that anyone is implying that

That seems more likely to yield reasonable results regarding the correlation between facial features and orientation

On the other hand, there are pitfalls of comparing averages between populations

A bit later in the paper, it says that they were able to get surprisingly high accuracy using only the face contour (75% for men and 63% for women). It looks like they were also able to get 75% accuracy for men using only the mouth, and using only the eyebrows

I got to the part where they talk about the signalling aspect. It says that human judges were able to guess orientation based on these pictures with the same accuracy as with pictures that were not sourced from dating sites in previous studies, which implies that "the facial images used in our study were not unusually revealing of sexual orientation(at least to humans)"

Overall, the paper's methodology seems pretty good, and the conclusions they drew seem reasonable based on the data presented.

@murgatroid99 so it's media taking a reasonable paper and making it into clickbait?

yeah, as always

@GodEmperorDune I don't think the Guardian article actually significantly distorted that paper's findings

@Ash I think that just wasn't part of the data set because they didn't have a sufficient volume of images

For the same reason, the study only involves white people, because there weren't enough non-white homosexual people in the data set for statistical significance

i do like a dash of racism in with my cishetsplaining

:D

The best splaining

The basic conclusion is that this program is significantly more accurate than humans at distinguishing between heterosexual and homosexual white people based solely on facial features

And that there is good evidence that genetics, hormone exposure, or some other physical process plays a role in sexuality, or at least is correlated with it

It's unfortunate, but that will only change when they can get significantly significant representation from those groups.

@Ash i've heard gatekeeper mentioned in these kinds of discussions... what does it mean specifically in these contexts?

@GodEmperorDune I would guess in this context it would be something like "You're not really gay, the classifier says you're straight"

@murgatroid99 ohhhhhhhh

yeah i didn't mean to provoke feels, but i understand the term now, thanks

@Ash I know asexual people, and I've heard about similar problems from them

@TimStone y u do dis

also why is blind creek beach over there if the fish have 3 eyes?

(and paper download)

Hahahahahahahahahahahaha holy shit

@Avery yeah i'm not going to retweet racist memes or opinion masquerading as "alternative fact"

It's also because people on the left are objectively better people

@Avery The responses to this make me a little sad. If you just characterize the other side as "racists", "mysoginists", etc. and don't think past that, you're perpetuating exactly that problem

@GodEmperorDune well same

This is a proven scientific fact proven with science

@murgatroid99 i listen to reasonable people on the right

I might RT opinions but vv rarely

@GodEmperorDune It seems like the whole point of that paper is that most people don't

actually some of my fave reporters lean right

being right or being left doesn't instantly make you terrible, but the majority of people you hear from on each group actually kinda is usually extreme

but i'm not going to relay infowars talking points, nor will i relay louise mensch junk

@Avery Relevant SMBC

@GodEmperorDune To be fair, I have literally no idea where one would qualify Louise Mensch in this context

@Unionhawk she spouts bullshit that seems to validate politically left dreams of impeachment

@murgatroid99 accurate

@murgatroid99 yes

@GodEmperorDune but like, nobody on the left believes anything she says ever, it's just people who think HRC was the best possible candidate

Which I guess is left by some standards

She doesn't translate as well as Infowars does as far as craziness=distance from center does, I guess

@Unionhawk she hits the craziness metric for sure

as distance from reality

marshal of the supreme court is equivalent to water turning frogs gay

Right

But everyone left of "HRC would have won" and people with a brain know she's full of shit

Whereas people left of "full fascism now" and people with a brain understand Infowars is full of shit

So I guess it actually does scan

I'm allowed to say that people who believe Infowars or Louise Mensch are dumb right

inb4 b&

ugh

That's a deviation from "in Ohio on Tuesday"

All of Florida is mandatory evac right

@Unionhawk who told you irma would hit Ohio?

@Unionhawk south florida for sure, i dunno north florida

plus mandatory evac is a mess for disabled people

@GodEmperorDune idk

@GodEmperorDune oh dear, it's going to go that much inland?

that's creepy.

They usually do

@Avery usual caveats with projections

yeah but like

But it breaks up by Tennessee or so usually

all projections I saw expected it to die off before it even got to like SC

but once you get to like 30 mph winds that's really not bad in areas used to full on hurricanes and tornados

We got remnants of Ike or something years ago that caused a major windstorm and school closures one year

I'm mostly worried about friends in NC but eh, they'll get hit with not too high speeds if they do get hit anyways so meh.

Yea @ori and @spu are there

Actually no @spu moved to Washington

like, at least quarter of my friends live in NC and that's kinda ironic but meh, there are a ton of stack people there too.

RIP @spugsley

i think one of the old mods, ravensomething, was in NC

yeah

ravendreams iirc

raven dreamer

speaking of which, I never asked what happened to them

@Avery It'll be no more than a category 2 by the time the eye is over Georgia.

I never ever saw them and they aren't that active but they're obviously alive according to post history but they still hold mod position

@Avery sometimes mods do things that are only visible to other mods

ofc

@GodEmperorDune Classic Florida

holy shit

5.6m? damn

@Ash :/ stay safe @ash's hubby

it's a pretty big MPH dropoff by the time it hits GA

its a prefectly reasonable thing to worry about though

Holy shit

Attention: don't bother checking to see if you're hacked in the Equifax thing, if you're in America you're hacked

we're in the worst timeline

Tbh I thought I was in bridge

I want my money back

rekt

The Bridge and This Is Fine start with the same letter

3confuse5me