1. The problems that Polaris are working on are exactly what Palantir software is made for.
2. Most other instances of this problem do not generate unequivocally good publicity.
To be absolutely clear, when I say "instances," I'm not referring to instances where Palantir is used. I'm referring to instances of the problem-type that Palantir solves.
Other examples of this problem and why they wouldn't generate unequivocally good publicity: counter-insurgency ("we shouldn't be involved in the middle east"), counterterrorism ("at the expense of civil liberties"), counterfraud ("rich people get away with fines after stealing millions anyways"), etc.
Palantir works with Polaris and NCMEC because they actually care (IMO), Polaris/NCMEC needs align precisely with Palantir capabilities, and they're a rare opportunity to be somewhat public about the company, if not the software.
`stopthelies comments brings the "unequivocally good" into question (maybe not to the general public, but certainly here in this thread), which makes this bit seem a bit handwavy if I state the circumstance like this: Even though we help enable organizations, whose methods are based on questionable ways of classifying what is going on to continue in order to attract funds/status/etc which can skew how resources are allocated and how certain populations would be negatively affected, this is still unquestionably a good thing to do. I wonder how many times "victims" are "helped" when "every country offers counselor services where they'll print them a new temporary visa and even a ticket back home"[0] and makes sure they're on the flight? We'll probably never know until everything isn't jumbled into "human trafficking" tag.
>Palantir works with Polaris and NCMEC because they actually care (IMO), Polaris/NCMEC needs align precisely with Palantir capabilities, and they're a rare opportunity to be somewhat public about the company, if not the software.
If I pass this through an altruism filter, the 2nd and 3rd remain. I can say sure to 2 because anyone with ability to generate such plots and access to the data could do that, but I think 3 is the most interesting. Can you elaborate on the benefits of this? The last bit makes me think someone could plausibly construct this as some kind of recruiting gambit for the company and not the unequivocally good cause (which of course would be a valid benefit, but only as far as others choice to perceive such as unequivocally good [i.e. act as a negative signal of sorts, kind of like whiteboard interviews for experienced devs]).
I'm not quite sure what evidence there is for this assertion that miscategorization is happening. The categories seem to be pretty straightforward. Unless you're accusing them of blatantly lying, I don't see the criticism here.
Examples directly from Polaris' "Facts" page: "The International Labour Organization estimates that there are 20.9 million victims of human trafficking globally... 68% of them are trapped in forced labor."
Maybe it should be more explicit that this means 14.2 million people are "slaves," but this is a matter of definition that Polaris and its peers have no control over. Other comments also go into why "slavery" isn't even an applicable term.
Further, this is not some charity-wide conspiracy of definition manipulation. In fact, on Polaris's "Labor Trafficking" page, you only ever see the 14.2 million figure: people forced into labor. I assume the most readily manipulated figure would be the last one on the page: "31% of undocumented Spanish-speaking migrant workers interviewed in San Diego County had experienced labor trafficking." If it is the case that NGO's are seizing situations like these to call every illegal immigrant who utilizes a coyote to cross the border a "victim of trafficking," then certainly it would surface in this figure.
View the study, though, and it looks like there is exactly zero effort put towards manipulating or even exploiting technical definitions. Section 2 of the paper: "Challenges in Trafficking Research." First problem: "Definitional problems." It then outlines which definitions compete for dominance, why, and then plainly states which one it uses. For the record, it can be found here (http://www.state.gov/documents/organization/10492.pdf, Sec 103 Paragraphs 8 and 9) and seems to make sense.
If these people are miscategorizing things, I simply can't see how. If they are lying, then I'd like to see evidence that that's the case. More evidence than an anonymous person claiming to be writing a thesis on the topic would be nice.
Again, people can construct whatever idea or explanation they'd like. If you really want to believe something is incapable of good then you can easily make even the most benign acts look insidious. The question is whether or not you can substantively provide evidence for what is otherwise paranoia.
You are free to your own opinions… one doesn't need to believe, and one shouldn't have to be swayed by stat inflation to address problems.
>I also don't think a comment on HN qualifies as bad publicity, and I haven't seen a substantive argument that presents the relationship that way.
I guess you have missed over the years bad press companies received because on threads of hn (Palantir may be above this affecting their bottom line though not all companies are as fortunate)… I've already asked for more sources from the commenter beyond the post they shared that pointed to some of the discrepancies in how people are counting, because I would actually be interested in seeing/collecting more accurate data from more nuanced definitions for myself (and it appears others in this thread would like to see that to).
>Maybe it should be more explicit that this means 14.2 million people are "slaves," but this is a matter of definition that Polaris and its peers have no control over. Other comments also go into why "slavery" isn't even an applicable term.
Sure, lets hand wave stat inflation and questionable definitions, and accept NGO's/governments/large corps have no control whatsoever compared to individuals as actors in a given society in how they aggregate data and generate their pretty plots.
>View the study, though, and it looks like there is exactly zero effort put towards manipulating or even exploiting technical definitions. Section 2 of the paper: "Challenges in Trafficking Research." First problem: "Definitional problems." It then outlines which definitions compete for dominance, why, and then plainly states which one it uses. For the record, it can be found here (http://www.state.gov/documents/organization/10492.pdf, Sec 103 Paragraphs 8 and 9) and seems to make sense.
Ok, the definition makes sense, so lets just accept it prima facie, and not question it or think about how we could improve upon it in certain ways that minimize even further how certain parties are negatively affected? How would people learn anything if we approached problems like that all the time? So just because one states how they're defining something, it absolves them from critique if people disagree and wish for something more nuanced? lol
>Again, people can construct whatever idea or explanation they'd like. If you really want to believe something is incapable of good then you can easily make even the most benign acts look insidious. The question is whether or not you can substantively provide evidence for what is otherwise paranoia.
There would have to be a component of fear or anxiety for this to be true, which if one has neither (i.e. Me), casting such as paranoia in order to bolster ones argument because one may not think an NGO or its corporate sponsors are being upfront or relying on "not being able to do anything about the definitions we benefit from at the expense of others to the degree of which we are not certain", then one must not "believe" (not sure why we have to believe when we can look at the definitional flaws that can negatively affect certain populations lives they may be seeking out) anything can be good is pretty weak, when noting how certain parties mere existence hinges on such (or at least their marketing pitches lol)… people are free to believe all the stat inflation they want, its not my time or after tax currency being sucked into the church lol
Do you work for Palantir or something? It would be good to know as I can not waste my time further hereticing in front of the choir.
I guess it comes down to this: you have the same question as I do. I'm curious to see what the actual evidence for the alleged manipulation is. Has the commenter who put forth the allegation re: stat inflation provided evidence for the claim?
The question of definition is fair in principle. I'd be glad to discuss whether or not the definition is fair. I'm of the opinion that it is, and if you're of the opinion that it's not then I really do want to know why.
Not sure what this discussion is, to be honest. You seem to be saying that we should be able to criticize these stats/categories/definitions. I agree with you, and then I'm asking you what your criticisms actually are. So far you're still pointing to "definitional flaws that can negatively affect certain populations." If there are such flaws with such consequences, that would be really problematic, I agree.
What are the definitional flaws, and how are they negatively affecting which populations?
This is not the area my paper covers, and I don't want speak about any particular NGO by name, but creating a hyperbolic tale of "sex slaves" makes money for NGO's. In reality, if it's in fact the nightmarish horror that image conjures, I'm pleased to let you know law enforcement, government and charity is abundant for these victims.
Victims are helped, perpetrators are punished. For example, in China, pimping can have you executed. Every country offers counselor services where they'll print them a new temporary visa and even a ticket back home.
Many choose not to, they'd rather go try to find a way to go through the immigration system and get citizenship. This is off topic of my paper, but being a female and having a child may earn the mother favor toward getting naturalized in many places (laws, department/ministry policies or judge discretion).
Regarding the different NGO actors, this is even farther away from my area. But sometimes NGO's are actually fronts for GO's. An example would be USA using NGO's for subversion in Russia (https://en.wikipedia.org/wiki/Russian_foreign_agent_law#Impl...). That said, way out of the area I'm collecting data on.
Regardless, this thesis is only a thesis. "Exposing" those who misrepresent NGO's isn't a career. What I want to tell you is that a lot of people pass on a humanistic message, when they're really hustling people.
If you do care about people, I think the best thing you can do is volunteer your time, skill and expertise. A good example is Médecins Sans Frontières.
I think the easy answer would just to be say no, but if politics certainly has the influence that you say it does in the incentive process of what is going on, then I'd assume their would non zero probability for a billionaire to swoop in and support you in telling the "truth" in a "careful and controlled" manor à la Intercept and the Snowden "leaks".
>This is not the area my paper covers…
What area's does your paper cover then?
>What I want to tell you is that a lot of people pass on a humanistic message, when they're really hustling people.
Oh I already know, it's always amusing to watch people take the bait and then ignore the actions of the same people out of mutual interest/ "loin des yeux, loin de coeur"/fear/ or can't be bothered to look deeper.
>If you do care about people, I think the best thing you can do is volunteer your time, skill and expertise. A good example is Médecins Sans Frontières.
One of my skills and expertise (mostly trying to apply things i'm playing around with to something that others might find useful) surrounds datamining/data exfil from walled gardens/etc, where one of most recent things I've done was building a tool[0] that let people search the hacking team emails more efficiently with LSA techniques so that people could use it to gain insight into the operations of international defense contractors in the realm.
If you have any suggestions on any possible methods of how one should go after looking for such info and to classify such as a signal of human trafficking that takes in consideration of the nuances that people are over looking, I'm all ears. I would be glad to take a stab at figuring out how to mine for such information if no one is doing it because I'm pretty curious to see how something like that would compare to what people are pushing now.
[0] https://github.com/cinquemb/hackedteam-email-index-mining