BornAPoorBlkChild posted:

https://twitter.com/Luckdecorcom/status/1738439329784217727?t=QbfVJTJRxMSA-YPeGWOMRQ&s=19


in what world do you give this to someone who you do NOT have absolute contempt for?

Salt Fish posted:

"joke" probations are the fiesta cat of our time

Bar Ran Dun posted:

some vehicles one can adjust the drivers side mirror far enough over to reflect it back at them. but not all vehicles unfortunately.

Paradoxish posted:

Drive thrus are actually the best place to do this because you'll occasionally get a reaction from the other driver, but also that reaction might eventually lead to you being shot

I do it anyway because it's very funny

Thorn Wishes Talon posted:

the number of wanna-be "revolutionaries" on this forum who would actually fight in a revolution is close to if not equal to zero

if their attitudes towards ukraine are any indication, they would in fact be the first ones to roll over and surrender to the big bad state, and they would justify it as minimizing needless death and bloodshed or whatever the gently caress

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

Salt Fish posted:

The best lovely job I had was dishwasher. It's a obscenely stressful, but you can steal so much food. Plus its a legit craft and it feels great when you can wash 1000 dishes in an hour and make it look easy. You get cut and burnt a lot and that kinda sucks, but overall its not bad. I fondly remember one month when I was super poor 100% of my calories were stolen from the restaurant. $0 food bill but I had to work 7 days a week. Idk, it was okay.

Pittsburgh Fentanyl Cloud posted:

Christmas is real

gradenko_2000 posted:

Are there really no regulations on headlights in the US?

What's stopping someone from mounting this on their car?

Canned Sunshine posted:

A world where people have a sense of humor?

gradenko_2000 posted:

Are there really no regulations on headlights in the US?

What's stopping someone from mounting this on their car?

LeeMajors posted:

you’d be surprised (or maybe not, actually) how many trucks and jeeps roll around SC with huge blinding LED light bars all night.

Thorn Wishes Talon posted:

the number of wanna-be "revolutionaries" on this forum who would actually fight in a revolution is close to if not equal to zero

if their attitudes towards ukraine are any indication, they would in fact be the first ones to roll over and surrender to the big bad state, and they would justify it as minimizing needless death and bloodshed or whatever the gently caress

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

(USER WAS PUT ON PROBATION FOR THIS POST)

biceps crimes posted:

the middle of nowhere town my family is from had a dollar general move in and it sucks rear end. nothing more american than dollar general swooping into a depressed area with almost nothing left and then sucking it all out

Paradoxish posted:

Dollar General's laser focus on areas that are either extremely poor or in terminal decline (or both) is equal parts horrifying and fascinating. It's like the chain store equivalent of a payday loan place opening up next door.

Morbus posted:

There are a lot of reasons for this, but part of it has to do with the way the currently popular image generators work. The short version is that image generation is computationally very expensive, so the generative work is basically done at extremely low resolution and then upscaled. This inherently produces "smooth" images.

The longer version is that the core task these networks are trained to perform is to progressively hallucinate plausible images from noise. Basically what you do is:

1.) Get a whole shitload of images, and add random amounts of gaussian noise to them. Sometimes just a tiny amount of noise, so it's just a slightly noisy version of the original image. Sometimes a lot of noise, so the image is basically entirely noise.

2.) Train a network to identify the noise that was added, given just the noisy images (not the originals). This network is rewarded just based on how close its prediction is to the actual noise pattern that was added.

3.) Now, give the network pure noise as an input. Take the "noise" it predicts you added, and subtract it from the original (pure noise) image. What you will get is some kind of vague, impressionistic shape or splotch or pattern*. Essentially this is the most plausible underlying "signal" that the network thought existed in your totally random pattern.

4.) Now add a fairly large but not "total" amount of noise to that vague pattern, and feed that back into your network. It will now hallucinate some additional details that seem plausible given the training data.

5.) Keep doing this with progressively smaller amounts of added noise, and the network will come up with some plausible looking random image.

This gives you an engine that can produce random but plausible images. If you go through essentially this same process, but with labeled images, and add some details, you can train a network that generates an image given some text descriptors, but the underlying image generation machinery is more or less the same as in this simple case.

What you end up with is a machine that is designed to hallucinate progressively more detailed images out of random noise, by removing what it thinks is noise, and thereby keeping what it is a plausible "signal". So, if you want a really detailed image, you need to run it through more "loops", which is expensive. In particular, running this process at anything approaching a normal image resolution is prohibitively expensive. So to make the whole scheme workable, the process is run at potato resolution (like 128x128 or something ridiculous) and various upscaling techniques are used. Since the whole machine is basically just dreaming up images and filling in the gaps based on incomplete information, it will tend to produce plasticky or smooth looking images in cases where it doesn't have enough training data to plausibly fill in the details.

The other problem is in how the utility functions are defined. Essentially if you reward the network for coming up with the most "likely" image, it will hedge its bets and produce a very "averaged-out" kind of image, with can contribute to "smoothness". If you reward the network more contrastively, that helps, but too much and it may go overboard and every image ends up looking like an LSD trip. There is also usually a lot of baggage behind the scenes trying to prevent the model from producing inappropriate content, which tends to constrain the ways a model can add detail. Ultimately these things are fine tuned by people, who are more likely to accept a cheap, inoffensive but boring model over one that produces more "correct" images but with additional costs and problems.

anyway that's my ted talk

*One question is that, if you give it pure noise as an input, shouldn't it recognize the entire signal as noise and then, if you subtract that, you just end up with a blank canvas? The reason it doesn't work out that way is partially due to the inherent statistical impossibility of discriminating real noise from signal, and partially due to the way the network is trained (i.e. on real images with added noise).

Paradoxish posted:

Dollar General's laser focus on areas that are either extremely poor or in terminal decline (or both) is equal parts horrifying and fascinating. It's like the chain store equivalent of a payday loan place opening up next door.

Fitzy Fitz posted:

When this shows up your town is dead

the popes toes posted:

Best poo poo job I had was as an orderly ("mental health worker") at a state psychiatric hospital on the 3-11pm shift. My ward was essentially the craziest of the crazy, they'd never leave, and it was their forever home - Austin State Hospital.

Job was essentially to keep order, make sure they didn't hurt themselves or each other, keep the place tidy and sit with them in a dayroom and try to keep them occupied. They were all batshit crazy, but the longer you worked there and got used to the craziness, they more "normal" it all was and you began to see them as people with personalities, intelligence and humor.

The real good part was after "lights out" you could study and do class work and chow down on free hospital food.

the popes toes posted:

Best poo poo job I had was as an orderly ("mental health worker") at a state psychiatric hospital on the 3-11pm shift. My ward was essentially the craziest of the crazy, they'd never leave, and it was their forever home - Austin State Hospital.

Job was essentially to keep order, make sure they didn't hurt themselves or each other, keep the place tidy and sit with them in a dayroom and try to keep them occupied. They were all batshit crazy, but the longer you worked there and got used to the craziness, they more "normal" it all was and you began to see them as people with personalities, intelligence and humor.

The real good part was after "lights out" you could study and do class work and chow down on free hospital food.

the popes toes posted:

Best poo poo job I had was as an orderly ("mental health worker") at a state psychiatric hospital on the 3-11pm shift. My ward was essentially the craziest of the crazy, they'd never leave, and it was their forever home - Austin State Hospital.

Job was essentially to keep order, make sure they didn't hurt themselves or each other, keep the place tidy and sit with them in a dayroom and try to keep them occupied. They were all batshit crazy, but the longer you worked there and got used to the craziness, they more "normal" it all was and you began to see them as people with personalities, intelligence and humor.

The real good part was after "lights out" you could study and do class work and chow down on free hospital food.

Real hurthling! posted:

i knew a guy with this job once
on the off chance its you, are your initials G.G. ?

gradenko_2000 posted:

Are there really no regulations on headlights in the US?

What's stopping someone from mounting this on their car?

biceps crimes posted:

the middle of nowhere town my family is from had a dollar general move in and it sucks rear end. nothing more american than dollar general swooping into a depressed area with almost nothing left and then sucking it all out

gradenko_2000 posted:

Are there really no regulations on headlights in the US?

coelomate posted:

isn’t dollar general where most of those lead-flavored applesauce pouches were sold?

silentsnack posted:

does it cause the super-wealthy to lose money, or somehow threaten to deprive them of income? does it undermine the authority of the ruling oligarchy or otherwise anger or inconvenience them? no? then it's cool.

Nitevision posted:

Danann posted:

https://twitter.com/NotMountainChen/status/1738487409472586002

Danann posted:

https://twitter.com/NotMountainChen/status/1738487409472586002

quote:

Last week, wearing nice clothes and excited to finally hang out in person with her new colleagues, my daughter baked some cookies to surprise her friends. But when she arrived at the office, no one was there. One was away, another said it was too cold to make it to the office, and a few others didn’t bother to explain why they didn’t show up.

In the same, beautiful, high-end building resides another well-known non-profit. It shares a kitchen with the one my daughter works for, so she knows what’s going on there as well. Out of 50 employees or so, she testifies that on a regular day, only about five or six are present. On the busiest days — about 20 make it to work.

I find my daughter’s experience extremely worrying. How can one build something with no real human interaction? How can one create the energy, the momentum, the friendships, the great work when everybody is staying at home, wearing their pyjamas all day? Aren’t we all Zoomed out by now?

quote:

Moreover, I’d bet that apart from the most disciplined workers, working from home makes us all less efficient since distractions from social media, text messages, private emails and more are much more tempting when no one is around.

quote:

Amir Barnea is an associate professor of finance at HEC Montréal and a freelance contributing columnist for the Star. Follow him on Twitter: @abarnea1.

Precambrian Video Games posted:

So are these twerps getting paid by increasingly desperate commercial REITs to churn out this RTO copypasta \