stolenmerch

I'm not sure you can just casually tack on "and the lack of masking". Massachusetts does NOT have a mask mandate and Louisiana DOES - for everyone regardless of vaccination. [1] Mass. has a 1+ vaccination rate of ~75% whereas Louisiana has a rate of ~48%.[2] We can talk about actual mask compliance all day but I'm putting my money on vaccines being the life saver here.

[1] https://www.usnews.com/news/best-states/articles/these-are-t...

[2] https://covidactnow.org/?s=22131139

This isn't totally surprising and has been studied previously with other respiratory illnesses. [1] There is also a correlation with severity of symptoms as shown in this paper and other cited sources. I'm not going to speculate too much, but one would hope that as severe cases decline, so too should these cases of "long covid".

[1] https://pubmed.ncbi.nlm.nih.gov/31718695/

Of all the web pages I have from 20+ years ago for things like directories of files or collections of documents, the ones that are broken now are the ones with assets split out. Their folders get clobbered or separated from their index files and everything broke during some data migration to a new hard drive or whatever. The ones with the css or js inline still largely work. Images usually aren't base64 encoded but these don't have many images anyway. I've always thought I should reconsider doing simple, single file applications more for things I want to archive into the distant future.

A better comparison would be if there was some salient feature surrounding my hands such as me holding an object. The object would be the equivalent to the jersey number. It's clearly not cropping on the thing that makes me feel bad but it did anyway. To avoid this, it has to be trained to take into account all possible human reactions. Basically, it has to know that these pixel values activate some social controversy and it's best to avoid. Not saying this isn't possible, but it seems absolutely fraught with peril and potentially more harmful than the original saliency, especially for social constructs with no clear consensus.

Training a ML model to understand how "most sensible, considerate humans" would act is anything but straightforward. I'm not even sure most people would even consider cropping an athlete at the jersey number regardless of race or gender - it just doesn't make sense yet the machine seemed to do it at the same rate for male vs. female. Retrofitting discrimination onto this result only once you learn the label of the data isn't particularly useful. We want to know how to make good non-harmful predictions in the future.

They are very clear about 'intentional' vs. 'unintentional' harm in their judging metric[1], so I get that I think. However, there is no possible way a saliency model could crop an image without someone reading bias into it. E.g. only cropping a jersey, only cropping a face, not cropping at all, etc. can all be signs of historical bias. At some point it just becomes a rorschach test for the reviewer. Not saying discrimination isn't happening or can't happen, just saying this very tenuous and weak connection to "historical bias" interpretation lacks a certain rigor in a bug bounty program or model optimization.

[1] https://hackerone.com/twitter-algorithmic-bias

"Although we find no evidence the saliency model explicitly encodes male gaze, in cases where the model crops out a woman’s head in favor of their body due to a jersey or lettering, the chosen crop still runs the risk of representational harm for women. In these instances, it is important to remember that users are unaware of the details of the saliency model and the cropping algorithm. Regardless of the underlying mechanism, when an image cropped to a woman’s body area is viewed on social media, due to the historical hyper-sexualization and objectification of women’s bodies, the image runs this risk of being instilled with the notion of male gaze."

"Point multipliers are applied for those harms that particularly affect marginalized communities since Twitter’s goal is to responsibly and equitably serve the public conversation."

I'm sure Twitter is just responding to the controversy from Fall of 2020 and doing due diligence to address the problem. However, how do you award a bounty for addressing "risk of representational harm" due to historical biases not inherent in the model? Genuine question here and one I'm always curious about. Seems difficult if not impossible.

Can confirm. At the height of Peppa, we had a long (and hilariously fun) conversation with our 4 year old about whether the word 'lever' was pronounced "LEE-ver" or "LEH-ver". He also likes to ask if we'd like to "have a go" and other little Britishisms.

Would love to see some blind study on what people actually find "uncanny" vs. what they think is uncanny because they are told it's a generated image. Maybe I don't fully understand the concept of the "uncanny valley" but simply saying This Looks Shopped isn't uncanniness.

Depends on if this proposed program is voluntary or involuntary. I suspect the folks to self-select to sit down with a random family to cook and share a meal may not identify with the demographics you point out.

Link to paper: https://advisory.kpmg.us/content/dam/advisory/en/pdfs/2021/y...

This work is based on one person's Master's thesis in Sustainability. I lack the tools to determine if her work actually does vindicate the original study or not. It's frightening if so, but I'd also like to hear other experts weigh in before I start prepping for the end times.

I work at a University with an top-tier, elite MFA program (not in film) that attracts star-eyed students from around the world. I've seen so many really talented folks chewed up and spit out by a pretty brutal system where there is not much besides enormous debt and creative bankruptcy on the other side of graduation. However, I've learned not to be a dick about squashing dreams. Most of them know the reality in the back of their minds but are so enamored with the idea of the MFA that they are compelled to move on it.

Naw, the "mandatory viewing" is pretty much all prestige television and blockbuster movies. Marvel is just the brand you are particularly irritated with. It makes sense because most other forms of entertainment have become too fragmented to have a meaningfully shared experience with anyone. Music, games, sports, arts are so niche and idiosyncratic now that a water cooler conversation is hard. New blockbuster TV and movies are still mainstream and accessible enough to break the ice with someone.

Not go all in on the Macromedia Director Lingo + perl cgi scripting stack. It was not the future of interactive media like I thought it was at the time.

Really? We were amazed by the amount and quality of kids content. It's a completely separate section, so it's not part of the main Kanopy app, but it has tons of movies and dramatized readings of children's books. But yeah, the first thing I thought when I read this was "Kanopy".

Am I correct in reading that these NG sites aren't accessible without a side DOM synced to canvas content?

I agree, but "critical thinking" itself needs to be monitored or the methods of fact checking and narrative questioning can become weaponized and pervert it into conspiracy. Danah Boyd has some decent talks about this phenomenon. https://www.edsurge.com/news/2018-03-07-danah-boyd-how-criti...

It's such a small thing but I can't stand having the output of a cell evaluate above the cell. I get that it's part of the Observable reactive environment, but it breaks what we're used to with the command line and other notebook applications.

I don't doubt their methodology or that a problem exists, especially for urban workers. They interviewed 1,502 American adults from 50 states over a two-week period. However, something kind of seemed off about this:

> "...about half of those without broadband say they can do everything they need to do online with their smartphone. ...If you don’t have [a computer with broadband], you’re not really plugged into the modern economy,” Rainie said. "...The data shows you’re not capable of being the kind of social, political, and economic actor that people who have broadband are able to be,” he added.

Perhaps if folks tell you they can do everything they need with their phone you should believe them instead of insisting they become they social, political, and economic actor you think they need to be. Believe it or not, some people still prefer to visit their bank in person and their potential jobs don't involve needing Zoom.

Completely agree with you about Vietnam reaching its military objectives. However, I only see a couple of commenters disagreeing with that, which is not "a lot". Only curious about this because I wonder how prevalent that viewpoint is versus the perception of that viewpoint being prevalent. Perhaps this is off-topic but it seems this disparity between perception is central to so much of America's hangups with this war 50 years later.

My guess is because they're not really an alternative because they don't produce the same numbing effects many people seek in ethanol consumption. Also, they are still illegal, stigmatized, and difficult to find in many places.

American states are still "places" and each had a different response and thus we can use it to measure effectiveness. Doesn't much matter if you are talking about Illinois or Sweden in this case.

This is demonstrably false. I'm not saying any state did particularly well, but there is no clear line between "states that tried to stay open" and "suffering both worse fatality rates and worse economic impacts".

https://wallethub.com/edu/states-coronavirus-restrictions/73...

https://www.statista.com/statistics/1109011/coronavirus-covi...

https://www.usnews.com/news/best-states/articles/unemploymen...

https://carsey.unh.edu/COVID-19-Economic-Impact-By-State#:~:...).

I'd check the link again because it answers most if not all of your questions. It weighs (with gradated strictness) mask usage, restaurant closures, school re-openings, large gathering restrictions, health checks, travel, lockdowns, and more. You may have some quibbles with the methodology, and that's fine, but it really doesn't look like "garbage data".

Also, there is no medical consensus on this array of restrictions and I don't think he claims to overturn anything. There is a robust and healthy debate among scientists and medical professionals on the effectiveness of school closures, lockdowns, temperature checks, etc. The scientific certainty of mask efficacy in some circumstances simply can't be a stand-in for every pandemic restriction.

I understand this is a polarizing topic because it maps closely to political leanings, but we should all try to be more open-minded to good faith efforts to understand the data whether it confirms our biases or not.

The data comes from healthdata.gov[1] and sources[2] at the U.S. Census Bureau, the U.S. Bureau of Labor Statistics, the Kaiser Family Foundation, Ballotpedia, Editorial Projects in Education, Centers for Disease Control and Prevention, National Restaurant Association, Littler Mendelson, Husch Blackwell and Ogletree Deakins.

How is this garbage data? Maybe it is, but I think you may need to do more work here than just calling him names. Everyone seems to feel very strongly about this on both sides, but I get the sense he made a good-faith effort with legitimate data. Perhaps you could do the same if you feel his conclusions were incorrect?

[1] https://healthdata.gov/dataset/COVID-19-Reported-Patient-Imp...

[2] https://wallethub.com/edu/states-coronavirus-restrictions/73...

This is getting a bit silly here, but for starters, people wear clothes regardless of a law so that's not a good comparison. Also, to put it very bluntly, I don't think the number of lives saved is worth the social costs I've already enumerated.

It's not the mask that does harm. It's the temptation by authorities to make blanket mandates that can easily encroach too far in an effort to cover as many scenarios as possible for legal purposes. It's not a conspiracy theory, it's just in their best interests to make it as easy as possible to interpret and enforce. It's also not at all cut-and-dried what a "public space" is, so that's a bitter argument that will be had. There is a lot of public space that is wide-open and sparsely travelled. Does this need a mandatory mask mandate? If we were all perfectly rational this would not be that difficult, but we're not.

Given current and recent events in the U.S., it's also probably best not to give the police even more reason to be able to stop and question folks for what is a very minor infraction. With something like mask mandates, it's a highly visible signal that must be obeyed by every single person, so it's prone to enforcement abuse. In general, it's just not worth it to get flu numbers down, particularly when we weren't really scared as a society about the flu before now. As always, if you're sick stay home or even wear a mask out. Laws to enforce this? Not so much.

> That was never the argument.

Yes it was. It's easy to forget what the actual messaging was in the early days. Originally we were all talking about not using up medical masks or N95 masks, which were meant to "protect the wearer from contact with droplets and sprays that may contain germs."[1] Then we all switched to wearing these homemade masks and the messaging turned to wearing them was important to stop the spread to other people. And it had to be done in every single circumstance.

> Constant debating over which situations were safe or not for going maskless would have harmed our overall responnse to the virus.

That is legitimately up for debate. One could make a good case that a noble lie to the public does more harm because it erodes trust in the institution that fibbed in order to "keep it simple". Part of the reason we were in this mask controversy mess is because the authorities were saying one thing and scientist were saying other things in many cases. Not always, but in some cases.

[1] https://www.mayoclinic.org/diseases-conditions/coronavirus/i...

Fine, then the claim that "If masks work, and you don’t wear one, you are potentially killing people and worsening the pandemic." is off-topic and irrelevant since this is no longer true.