gvisor tries to be a complete kernel in userland we are not trying to. We will consciously choose never to try and support multi-proess env in the sandbox. The idea is there are enough people running single process containers and they can benefit from a lighter more secure runtime. This solution will not try to replace the kernel. For example the python tests we run for https to some website ends up runnign implementing only 60 syscalls not 350. i expect to add another 10-20 for support typescript but this will always be strictly single process.Plus the performance overhead of gvisor is substantial 2-10us ( me reading internet) for the system i am implemeting on the hot path it is less than 1us. Plus there is always the density story my shim currently is 4KB the python runtime is shared through memfd. I am working on a demo showing i can run 1000 vm on 512 MB ram each launching in under 30msec.
Remember this will never replace or be able to handle generic mutli-process sandboxes this is targeted only at single process env where we can make lots of simplifying assumptions
Its a fantastic project, but seems almost inactive now. I have a tiny PR pending for weeks, even reviewed, but not merged. I have another patch I have not submitted as I want to first navigate the earlier PR to completion. Both were bugs that bit me, and I ended up wasting quite a lot of time trying to find it:
1. go:embed supports "all:<pattern" while tinygo silently ignore it. I ripped my hair out trying to figure out why my files were not showing up in embedfs. PR pending.
2. go allows setting some global vars at the build cli (think build version/tag etc). In the code, one can define a default, and then the value provided (if any) on the cli can override it at build time. Tinygo fails to override the value at build-time, silently, if a default value is provided for the var in code. This PR I have not submitted yet, as its more intrusive.
I hope it picks up steam again soon. I love using go for embedded and CF worker use and tinygo makes both of these use cases much more viable than regular go. Honestly, I hope tinygo can be rolled over into the main go toolchain as "target arch".
Honestly, I am vastly more interested in how they built the model: its not backprop-trained, it was done manually. And how did that bug find its way it?
I long time ago, I was operating a "social network" which allowed image uploads. (India local, didn't amount to much as was shutdown.).
Immediately at launch, we started having a huge amount of (image) pron being uploaded into the pages. We put in some rate limits etc, but did not want to put any major restrictions of user signups etc as that would hurt signup figures (important to the ceo!).
We already had some content review people thru a temp agency on site, so we checked with them and they were fine doing this manual filtering of these images for us. All young (early 20s) women. While my team built a quick "dashboard" for them to be able to do this image filtering quickly and conveniently, I had a detailed conversation with them as I was very concerned about having them review this kinda stuff for 8 hours a day, 5 days a week. Truly nasty stuff.
They were _perfectly_ clear that they had no issues with it, and they told me in so many words to not give it a second thought and to let them get back to work.
It was a surprise, but it was a point of realization: there are much worse things they could be doing. And looking a porn has shock value only the first time. I was under-estimating these women and assuming they were some "snowflakes" who could not deal with something this silly and non-threatening.
It's good that you warned them in advance, but I don't think "there are much worse things they could be doing" really is a reason to discard those who are hurt by this as "snowflakes" and the content they're exposed to as "silly and non-threatening."
From what you're saying, there's also a big difference between what you were dealing with and what this article is talking about: in your case it was only still images, and only "regular" porn (and I assume they probably had to monitor every image posted if it was "a long time ago" so overall they were not exposed only to that content); in this case they're reviewing videos content including not only porn but also violence and sexual abuse, that was pre-flagged by an AI so with way less rule-abiding content mixed in.
People butcher livestock for pay on a daily basis, rendering a live animal dead and cutting it to bits, on a massive scale to provide the food that we eat. Rescue and medical personnel deal with the injured, ill, and dying on a daily basis on a massive scale. Merely watching videos, even of disturbing material, doesn't even come close to being as bad as those and some other professions.
> Merely watching videos, even of disturbing material, doesn't even come close to being as bad as those and some other professions.
The research and reporting, which looks at actual people's experiences says otherwise. This issue has been coming up for years.
People live through war, murder, torture, rape, etc. We can always find something worse but that doesn't make the current situation better. Human experience of pain and trauma isn't scaled relative to the worst possible pain and trauma.
I think watching children being butchered or prisoners being tortured to death could be equally or more traumatic. Killing in processing animals is at least socially acceptable and (arguably) necessary for survival.
As I've said, images can be horrific, and some people can be traumatized by them. That must not be dismissed.
However, It is also important to carefully and properly quantify these things and not sensationalize. You've linked a 50+ minute documentary, without comment, that seems to prove that one person hired to curate content can become traumatized by that process. I can't be certain that is what it is about, because I will not waste time watching documentaries (the vast majority of which are outright propaganda or incredibly biased, while pretending to be objective), but still, I've no doubt the general claim is true, since I never claimed or believed otherwise.
But you've not provided meaningful statistical or scientific evidence properly quantifying such harms in general.
You stated an opinion but i gave you hard data by pointing you to a documentary which you are refusing to watch. You have not even read the description for the video which gives you a ToC with timestamps to jump to for the major themes. I can point you to lots of relevant data on the web but the onus of educating yourself is on you.
If you had just spent 5-10mins watching excerpts from above timestamps you would have seen interviews with workers doing such jobs from rural America/Canada/Spain/France/etc., sleazeball CEOs of sleazy companies taking these sort of contracts from Facebook/etc. and then farming it off to poorer parts of the World in Asia/Africa/South America/etc., psychologists who study and warn of the very real dangers of such a job etc.
So before asking for more data, start with the one i have already given you. The psychological harm caused by repeated exposure to graphic imagery is well studied; AI image annotation is one subset of that.
"Some workers report nightmares in which the violent images they reviewed replay in gruesome loops. Others experience intrusive flashbacks while riding the bus or shopping for groceries. Over time, many describe a numbing of their emotions-a flattening of joy, sorrow, or empathy-because the only way to cope is to feel nothing at all."
and bla bla bla. "Some", "Others", "many"... Wikipedia itself couldn't generate a better example of "weasel words". Like I said, we all know some people are traumatized, but all that matters is the amount. What if some is less than 1%? Less than 5%? More than 50%? The answer matters, but you are not providing answers to this.
Also, learn to basic science. Anecdotes are not data.
No, I am not watching a documentary for a collection of anecdotes, I clearly explained why documentaries don't count as serious sources of info when trying to accurately quantify things.
EDIT: And if you really know that traumatization from images on computer screens is "well-studied", you can surely link to one to three of such studies, rather than lame documentaries telling cherry-picked sob-stories.
Your opinions have no basis in facts; talking about quantitative statistics without having any idea of the raw data is ignorance. The documentary/article point you to such data. The article in particular has links to others including the book Ghost Work - https://ghostwork.info/ which contains lots of data. AI content annotation falls under this umbrella.
I had already mentioned that graphic imagery causing psychological harm is well studied. Psychologists call it Secondary Trauma with symptoms similar to PTSD (which you have helpfully noted above) - https://www.ptsduk.org/secondary-trauma/?ref=thebrink.me
In fact, Facebook was taken to court over this, forced to acknowledge the harm done and paid out a hefty amount;
A simple google search would have given you any number of articles/papers on the subject. But instead of educating yourself, you are merely asking to be spoon-fed.
Nevertheless, start here (two broad classes of graphic imagery);
I really don't think it's fair to minimize someone's struggles just because their situation could be worse. Is only the most miserable person on the planet (by what metric anyway?) allowed to complain about their condition?
I also don't think it's fair to exploit people who are in terrible situations by pushing jobs we don't want onto them, pay them a handful of crumbs, and then say they should be happy with what they get because their neighbor who does another job gets half a handful of crumbs.
The women in this situation aren’t complaining. Very much the opposite.
Why the compulsion to paternalistically ‘protect’ everyone even to the point of making them unemployed? I assume they weighed their options and decided this was the best one. It sounds like you want to stop them from doing that?
You're misrepresenting both what the article says and what I wrote.
The article explicitly mentions that the jobs aren't clearly labeled so they couldn't weigh their options beforehand, that concerns raised by the workers are being dismissed by management, and that several workers have developed mental health problems.
I'm not arguing that these women should be "protected" by taking their jobs away or that they can't make their own choices; of course they're weighing their options and deciding accordingly (the article even mentions that some of them decided to leave). But it's not unreasonable to critique a system where the only choices they have are all horrible in (often more than) one way or another.
I actually had a conversation about this with my mom. We were talking about the hotel cleaners in Dubai walking around with toothbrushes to clean the shower which seemed mildly ridiculous to our European eyes.
But we came to the realisation that these folks were probably happy that they could send money back to their villages. And we left a nice tip.
I don't question your good faith. You seem to have fallen into a trap of many of good faith (and those of not-so-good faith):
For ~ the first half of the 20th century, the leading scholarly theory on US slavery was that the slaves were happy living in a civilized land, etc. You'll see the same claim about many things.
I think the trap has two sufficient components: First, it lacks empirical observations - actual fact that tethers us to reality - and therefore is very prone to drift far from the ground truth. Second, it's a much more comfortable worldview for us, and we tend to adopt such worldviews until compelled to do otherwise.
Just because there is an upside for them - sending money home - doesn't make them happy or make it good or ok. As an extreme example, some underage people go into prostitution because they need the food and shelter to survive - the fact that they get those benefits doesn't make them happy or make it good or ok.
Wow, what a way to conflate consenting adults making rational economic choices (which, btw, the finger wagging in the article and here doesn’t actually PROVIDE BETTER ECONOMIC CHOICES) with…. child prostitution?
The reality is that I’m not seeing anyone proposing actually making the situation better, or addressing why people might consider these jobs better than their alternatives.
Rather people trying to shut down these jobs - and defacto pushing other people into what those people clearly seem to consider to be worse alternatives?
Having actually seen up close and personal the alternatives many of these people are facing, while I think everyone would of course prefer the nice comfy joys of an office programming job, no one here seems to be offering those people those jobs are they? In fact, people are scrambling to keep those jobs.
Instead, those folks would end up taking a similar job elsewhere, or even more fun - something likely far worse. Which is why they are applying for those jobs in the first place.
Can we make the jobs a bit better? Likely, and we should.
Does it change the nature of those jobs? Or make the alternatives better? Not a bit.
Is this attached to the wrong comment? We weren't talking about the OP.
> I’m not seeing anyone proposing actually making the situation better, or addressing why people might consider these jobs better than their alternatives.
My statement is not putting the failure in the individuals who are accepting the work as a means of survival.
My original statement stands, there is plenty of productive work to be done in the world, but industry is funneling labor into man-made horrors for bottom dollar wages
Wouldn't it be better to give women more choice in general and the option of therapy if they take on a job with a risk of PTSD? Dismissing the concerns of the women actually doing the work isn't making things any better.
That not only doesn’t answer my question, but might as well be ‘have you stopped beating your wife yet’.
And how have you done re: ongoing jobs? The thing that was the actual topic here?
The reality is that if there were better jobs available for folks relative to number of available employees, the employer wouldn’t be able to find the people to do the crap jobs we’re talking about, and would have to make them better to get anyone to do them.
Or result to actual (or just de-facto) slavery anyway, which does happen in some places.
It’s a continuum.
Actual living conditions in India, in particular, are pretty brutal for most people. But certainly not just there!
While one can blame corporations, the most blame lands on the Indian government(s). Decades and decades of corrupt local, state, and central governance has led to dire poverty and high levels of unemployment. The current and past leaders have had no care to fix it, and it’s only getting worse. Their incompetence is what creates these kinds of jobs as alternatives to abject poverty and death.
As an Indian living in India's Silicon Valley, I am calling BS on this.
"All young (early 20s) women" doing explicit content manual filtering is not normal. No Indian Family would allow their daughters (and sons) to take up such a job if they knew about it.
As the article itself points out, the workers are drawn in by non-explicit/non-pron image annotation and then switched to psychologically harmful work when they cannot leave due to job-needs/contracts-enforcement.
This is pure bait-and-switch scam which has severe psychological effects on the women (and men) involved and should not be blithely hand-waved away.
For a detailed understanding of the psychological harm of explicit image annotation see this documentary (not specific to India) The "Modern Day Slaves" Of The AI Tech World. Undercover reporter does annotation of explicit images for AI but is so traumatized by what he sees that he quits in two weeks. - https://www.youtube.com/watch?v=VPSZFUiElls
As a RA-ship student in the US, I still had taxed withheld on my meager grad student "salary", and was required to file a 1040NR-EZ. I obviously did not want to spent 50$ at H&R block, to claim like 200$ back. Fortunately the local library had sessions where you could grab a blank form (the library had boxes full of them), and someone walked a bunch of us (i think there were like 40 in the room), to filling it out. The session mailer had already listed the stuff we were supposed to bring with us at the session. Even if someone forgot, they could leave a place holder and take care of it later.
The session was an hour long, and I was done with my return by the end of the hour. I dropped it off in the mail on my way back from the library and that was the end of it. From the subsequent year onwards, a pre-printed form arrived in the mail with like 90% of the stuff prefilled, and it took like 5 minutes to fill out the rest and drop it in the post box.
I honestly didn't think it could have been any easier - of course, not having taxes withheld from a far-below-minimum wage salary would have been nicer.
When I lived in the US(5 yrs), early 2000s, from my second year onwards, I used to receive a pre-filled 1040NR-EZ with my W2 info already printed/filled-in on it. Typically, I would just add a deduction, and mailed it back. Does that program not exist? Or was it only for NR?
I have always counted to 20 on one hand. even as a kid. base, lower joint, upper joint, top. times 5 - including the thumb: my motor memory is trained so that i switch seamlessly from keeping the curse on top of the finger using my thumb, and then, once i cross 16, switch to using the index finger to "cursor" the thumb.
Same here. I have always counted 20 on one hand, so 40 with both. That's how my parents taught me to count when I was little. I used this method so often as a kid that, even though I don't count like this anymore, every number up to 40 still has its own place on my fingers.
It was only as an adult that I realised nobody around me counted this way. You are the first person I have found who talked about this method, so I am glad to find this comment of yours.
I asked this question on another post and was downvoted, trying again: don't we lose the "contextualization" that LLM embeddings do (embedding on Token X contains not just information about X, but also of all tokens that came before X in the context, causing different embedding for "flies" in "time flies like an arrow" vs "fruit flies like a banana")?
The image embeddings, as I currently understand, are just pixel values of a block of pixels.
But does this not miss the "context" that the embeddings of the text tokens carry? An LLM embedding of a text token has a compressed version of the entire preceding set of tokens that came before it in the context. While the image embeddings are just representations of pixel values.
Sort of at the level of word2vec, where the representation of "flies" in "fruit flies like a banana" vs "time flies like an arrow" would be the same.
Isn't this "as intended" in the westminster-style system? The govt is formed by MPs from the majority party (or alliance). By definition they MUST be able to pass ALL money bills, which only require a simple majority. Any failure to pass a money bill is equivalent to the govt no longer holding a majority support in parliament. And that means either the king/president/govgen invites someone else from the current parliament who they have good reason to believe DOES (potentially) have support of majority of the parliament, or dissolve the parliament and call fresh elections if there is no such majority.
I am not quite sure why an action with such a clear established precedent be considered foreign interference? or was it the case that there WAS a suitable candidate with a possible majority but they were NOT invited by the govgen to try and win a trust vote in parliament?
It was very much an edge case, with one of Whitlam's senators on leave and recent changes to territory rules giving additional senators to the opposition party (as I recall ...) the ability to block supply appeared suddenly out of the blue.
Whitlam did move to call an election (rather than be sacked) which likely would have removed the blocked supply threat as he was at the time an extremely popular PM in Australia (loved by the common masses, despised by many elites) .. and when attending the Queens Repreresentative (the Governor General) to advise about calling an election .. he was removed by the G-G.
Strictly speaking the "as intended" outcome should have been to resolve a looming (not yet happened) supply crisis by allowing the people of Australia to vote, instead the government of the day (Whitlam's) was removed on a technical reading against the spirit of intended resolution.
There's a peer comment here that linked to a 2020 article on the finally released royal correspondance that's worth a read. The US influence angle has merit also, they had weight in the game for sure, how much and whether it tipped the balance is debatable.
Literally reams of contraversay here, the G-G acted autonomously and likely to save his own neck as Whitlam intended to replace the G-G, additionally many outside powers (the UK and the US) were whispering in the ears of those with levers to pull seeking to dump Whitlam; he was returning real power to the people, providing socialised health and education to the masses, asking questions about the role of secret American bases on AU soil, etc.
