I understand and agree with frustrations around non-US citizens/permanent residents being ineligible for prize money. None of Kaggle, DHS or the community wants this restriction.
\n\nThe reason for the restriction is the America COMPETES Act. DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.
\n\nFrom Kaggle's perspective, we had a choice: host this competition with a geo restriction or not host at all. This was a tough choice. Hosting the competition meant running a two-speed competition, where some are eligible for prize money while others are not. Doing this violates our desire for meritocracy, where we offer equal opportunities to every community member regardless of education, background and (of course) country.
\n\nIf we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning. Not exposing this dataset to our community meant depriving our community, and the machine learning world more generally, of a chance to push forward the science with a novel and challenging dataset.
\n\nWe decided that it was better to make the dataset available. Many in this forum have said this was a mistake. This was a challenging issue; I hope this at least gives you a window into our thinking.
\n\nAnthony
", "messages": [ { "id": 200775, "postDate": "2017-07-09T08:42:22.337Z", "content": "Anthony, according to COMPETES ACT an american entity can receive the prize. So Kaggle could be that entity that receives the prize and later pays to non American people.
\n\nIt's as simple as that.
", "rawMarkdown": "Anthony, according to COMPETES ACT an american entity can receive the prize. So Kaggle could be that entity that receives the prize and later pays to non American people. \n\nIt's as simple as that.", "votes": 63, "replies": [ { "id": 200952, "postDate": "2017-07-10T01:47:36.793Z", "rawMarkdown": "", "votes": 1, "isDeleted": true }, { "id": 205007, "postDate": "2017-07-20T12:29:04.537Z", "content": "Why such a good suggestion by ironbar is being ignored.
", "rawMarkdown": "Why such a good suggestion by ironbar is being ignored.", "votes": 3 }, { "id": 208725, "postDate": "2017-07-30T19:17:28.133Z", "content": "are you certain that Kaggle is an American entity?
", "rawMarkdown": "are you certain that Kaggle is an American entity?", "votes": 1 }, { "id": 209853, "postDate": "2017-08-03T15:55:58.077Z", "content": "If it were that simple, the COMPETES act would have been formulated differently just to make sure that it weren't that easy to circumvent..
\n// Yep this statement loops on itself :P
For Kaggle to do that, Kaggle will need employment permission for each non-USA participant for this challenge.
", "rawMarkdown": "For Kaggle to do that, Kaggle will need employment permission for each non-USA participant for this challenge." }, { "id": 229036, "postDate": "2017-10-08T15:34:33.750Z", "content": "If a solution by a non-USA participant is better than USA participants, Kaggle may purchase the solution from that participant, \"bottle\" it, and sell it. A win-win situation!
", "rawMarkdown": "If a solution by a non-USA participant is better than USA participants, Kaggle may purchase the solution from that participant, \"bottle\" it, and sell it. A win-win situation!" } ] }, { "id": 200691, "postDate": "2017-07-08T22:56:56.540Z", "content": "Anthony,\nYour (only) argument is that the “availability of datasets” to push forward ML. Yet, competition rules say:
\n\n\n\n\nThese images are for competition participants only. You should not share the images publicly, plan to use them in future research, or blog about them. You are also asked to destroy the data once the competition is over.
\n
Care to elaborate?
\n\nI believe you are bluffing here – I believe the sole reason is Kaggle's financial gain.
", "rawMarkdown": "Anthony,\nYour (only) argument is that the “availability of datasets” to push forward ML. Yet, competition rules say:\n\n>These images are for competition participants only. You should not share the images publicly, plan to use them in future research, or blog about them. You are also asked to destroy the data once the competition is over.\n\nCare to elaborate? \n\nI believe you are bluffing here – I believe the sole reason is Kaggle's financial gain. \n", "votes": 53, "replies": [ { "id": 204099, "postDate": "2017-07-17T15:39:22.493Z", "content": "Wow that's something that I didn't notice, I'm glad you pointed it out.
", "rawMarkdown": "Wow that's something that I didn't notice, I'm glad you pointed it out.", "votes": 1 }, { "id": 206681, "postDate": "2017-07-25T02:46:45.790Z", "rawMarkdown": "", "votes": 1 } ] }, { "id": 201953, "postDate": "2017-07-11T16:16:15.973Z", "content": "\n\n\nIf we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning.
\n
If so, make this dataset really available for us to learn. Currently it is available at terms like \"give us everything as community, we give nothing (to most of you)\". Allow unrestricted usage of data and we done (maybe, I couldn't talk for everybody here).
", "rawMarkdown": "> If we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning. \n\nIf so, make this dataset really available for us to learn. Currently it is available at terms like \"give us everything as community, we give nothing (to most of you)\". Allow unrestricted usage of data and we done (maybe, I couldn't talk for everybody here).", "votes": 13 }, { "id": 200588, "postDate": "2017-07-08T15:52:56.197Z", "content": "I understand and agree with frustrations around non-US citizens/permanent residents being ineligible for prize money. None of Kaggle, DHS or the community wants this restriction.
\n\nThe reason for the restriction is the America COMPETES Act. DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.
\n\nFrom Kaggle's perspective, we had a choice: host this competition with a geo restriction or not host at all. This was a tough choice. Hosting the competition meant running a two-speed competition, where some are eligible for prize money while others are not. Doing this violates our desire for meritocracy, where we offer equal opportunities to every community member regardless of education, background and (of course) country.
\n\nIf we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning. Not exposing this dataset to our community meant depriving our community, and the machine learning world more generally, of a chance to push forward the science with a novel and challenging dataset.
\n\nWe decided that it was better to make the dataset available. Many in this forum have said this was a mistake. This was a challenging issue; I hope this at least gives you a window into our thinking.
\n\nAnthony
", "rawMarkdown": "I understand and agree with frustrations around non-US citizens/permanent residents being ineligible for prize money. None of Kaggle, DHS or the community wants this restriction.\n\nThe reason for the restriction is the America COMPETES Act. DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.\n\nFrom Kaggle's perspective, we had a choice: host this competition with a geo restriction or not host at all. This was a tough choice. Hosting the competition meant running a two-speed competition, where some are eligible for prize money while others are not. Doing this violates our desire for meritocracy, where we offer equal opportunities to every community member regardless of education, background and (of course) country.\n\nIf we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning. Not exposing this dataset to our community meant depriving our community, and the machine learning world more generally, of a chance to push forward the science with a novel and challenging dataset.\n\nWe decided that it was better to make the dataset available. Many in this forum have said this was a mistake. This was a challenging issue; I hope this at least gives you a window into our thinking.\n\nAnthony", "votes": 31 }, { "id": 202382, "postDate": "2017-07-12T12:02:25.300Z", "content": "Anthony, a solution adopted by some sport competitions could solve the misunderstanding. To clarify, let us take the example of a city hosting an international marathon and the mayor wants to motivate local runners to participate. The fastest runner wins the general prize from the global sponsors and the fastest runner from that city wins an additional prize.\nThis strategy maintains the global competition feature and allows governments, institutions and companies to provide a local stimulus following the specific criteria they want and respecting legal restrictions.\nIn this competition, the best model(s) could receive a prize from Kaggle and the best competitors who meet the entity criteria (America COMPETES Act of 2010) are qualified for an additional prize paid by U.S. Transportation Security Administration (TSA).\nBest Regards,\nEduardo.
", "rawMarkdown": "Anthony, a solution adopted by some sport competitions could solve the misunderstanding. To clarify, let us take the example of a city hosting an international marathon and the mayor wants to motivate local runners to participate. The fastest runner wins the general prize from the global sponsors and the fastest runner from that city wins an additional prize.\nThis strategy maintains the global competition feature and allows governments, institutions and companies to provide a local stimulus following the specific criteria they want and respecting legal restrictions.\nIn this competition, the best model(s) could receive a prize from Kaggle and the best competitors who meet the entity criteria (America COMPETES Act of 2010) are qualified for an additional prize paid by U.S. Transportation Security Administration (TSA).\nBest Regards,\nEduardo.", "votes": 7, "replies": [ { "id": 202531, "postDate": "2017-07-12T18:08:59.003Z", "content": "good points!\nMore extreme: no cash prizes, everybody can participate, everybody receives Kaggle points.
", "rawMarkdown": "good points!\nMore extreme: no cash prizes, everybody can participate, everybody receives Kaggle points." } ] }, { "id": 220913, "postDate": "2017-09-13T13:29:33.637Z", "content": "Could you make it if a non-American wins they can donate their winnings to a charity?
", "rawMarkdown": "Could you make it if a non-American wins they can donate their winnings to a charity?", "votes": 4 }, { "id": 203211, "postDate": "2017-07-14T10:46:46.453Z", "content": "Can we use this dataset for academic purposes in the future?
", "rawMarkdown": "Can we use this dataset for academic purposes in the future?", "votes": 4 }, { "id": 200597, "postDate": "2017-07-08T16:38:20.100Z", "content": "All the controversy behind this competition / dataset / DHS data provider reminds of a Netflix series titled 'The Machine'. Spoiler alert - throughout the series, the creator of this over-arching, all-seeing AI battles with the knowledge that his creation explicitly is being used to kill people without due course of law (and in some instances, it even directly kills people itself). He rightfully understood that if he didn't do it, someone else more sinister would. His desire in building it was to use technology to stop other people from doing bad things... and for the most part, was very successful in doing this---depending on who got to define who the 'bad' people were, of course. Throughout the series, we sometimes learned that the people we felt were bad actually weren't, and it was in fact the good guys who were very bad, or vice versa.
\n\nAnyhow, I don't really have much to add to this discussion except my belief that as people... as data hackers, explorers, engineers, scientists, or whatever, it really is our moral responsibility to see through to the end of our actions. By that I mean, sure, it might be interesting data and in interesting problem; but just looking one step ahead at the immediate goal of wining a contest, acquiring some fame, coming into some money, etc. when the technology we're working with is exponential in nature, and the organizations we develop the technology for are national, and the implications of our work undoubtedly will effect hundreds of millions of people, kinda is deserving of at least an introspective discussion, if not a public one. When added on top of that is the culture conflict with the contest host, iono... =\\
", "rawMarkdown": "All the controversy behind this competition / dataset / DHS data provider reminds of a Netflix series titled 'The Machine'. Spoiler alert - throughout the series, the creator of this over-arching, all-seeing AI battles with the knowledge that his creation explicitly is being used to kill people without due course of law (and in some instances, it even directly kills people itself). He rightfully understood that if he didn't do it, someone else more sinister would. His desire in building it was to use technology to stop other people from doing bad things... and for the most part, was very successful in doing this---depending on who got to define who the 'bad' people were, of course. Throughout the series, we sometimes learned that the people we felt were bad actually weren't, and it was in fact the good guys who were *very* bad, or vice versa.\n\nAnyhow, I don't really have much to add to this discussion except my belief that as people... as data hackers, explorers, engineers, scientists, or whatever, it really is our moral responsibility to see through to the end of our actions. By that I mean, sure, it might be interesting data and in interesting problem; but just looking one step ahead at the immediate goal of wining a contest, acquiring some fame, coming into some money, etc. when the technology we're working with is exponential in nature, and the organizations we develop the technology for are national, and the implications of our work undoubtedly will effect hundreds of millions of people, kinda is deserving of at least an introspective discussion, if not a public one. When added on top of that is the culture conflict with the contest host, iono... =\\", "votes": 3 }, { "id": 203555, "postDate": "2017-07-15T13:40:30.110Z", "content": "One workaround for non-US people can be to partner with a US person whom you can trust, let that person get the prize money and later you can share among yourselves.
", "rawMarkdown": "One workaround for non-US people can be to partner with a US person whom you can trust, let that person get the prize money and later you can share among yourselves.", "votes": 1, "replies": [ { "id": 207778, "postDate": "2017-07-27T12:22:16.817Z", "content": "\"Members of the Kaggle community who are not United States Citizens or legal permanent residents at the time of entry are allowed to participate in the Competition but are not eligible to win prizes. If a team has one or more members who are not prize eligible, then the entire team is not prize eligible.\"
", "rawMarkdown": "\"Members of the Kaggle community who are not United States Citizens or legal permanent residents at the time of entry are allowed to participate in the Competition but are not eligible to win prizes. If a team has one or more members who are not prize eligible, then the entire team is not prize eligible.\"", "votes": 1 }, { "id": 244450, "postDate": "2017-11-16T10:30:01.777Z", "content": "I think partnering could be done implicitly; you work together but don't form an official team on Kaggle.\nIt's like forming a team with your family or friends but use one personal profile to post solutions. I'm not sure if that would be a violation of rules though.
", "rawMarkdown": "I think partnering could be done implicitly; you work together but don't form an official team on Kaggle.\nIt's like forming a team with your family or friends but use one personal profile to post solutions. I'm not sure if that would be a violation of rules though." } ] }, { "id": 207781, "postDate": "2017-07-27T12:31:39.717Z", "content": "\"DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.\" Perhaps you should ask Alphabet Inc legal team to look into it, maybe they will have more luck.
", "rawMarkdown": "\"DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.\" Perhaps you should ask Alphabet Inc legal team to look into it, maybe they will have more luck." }, { "id": 657927, "postDate": "2019-10-25T14:17:32.460Z", "content": "well done
", "rawMarkdown": "well done", "votes": -1 }, { "id": 209725, "postDate": "2017-08-03T07:05:06.380Z", "content": "It is a well known fact that Google \"lacks\" money and cannot subsidize a prize in a non discriminatory way like \"should a non US citizen member of our community win the competition, we'd make sure he/she'd still happy to win\"... just saying, on the other hand, it might be a bit more complicated than that.
\n\nThis being said, discriminatory compromises even for the \"sake of a dataset\" have only taught one thing: it is ok to discriminate.
", "rawMarkdown": "It is a well known fact that Google \"lacks\" money and cannot subsidize a prize in a non discriminatory way like \"should a non US citizen member of our community win the competition, we'd make sure he/she'd still happy to win\"... just saying, on the other hand, it might be a bit more complicated than that.\n\nThis being said, discriminatory compromises even for the \"sake of a dataset\" have only taught one thing: it is ok to discriminate.", "votes": -1 }, { "id": 252567, "postDate": "2017-12-03T08:48:10.707Z", "content": "I have informations from the officers that are monitoring this individual Anthony Goldbloom for creating a gamification game called Kaglle , that nearly 99% of the competitions are viciated by information leks and the solutions trucate , the winners are well know even before the competition is hosted .
", "rawMarkdown": "I have informations from the officers that are monitoring this individual Anthony Goldbloom for creating a gamification game called Kaglle , that nearly 99% of the competitions are viciated by information leks and the solutions trucate , the winners are well know even before the competition is hosted .", "votes": -6 }, { "id": 204188, "postDate": "2017-07-17T20:53:22.100Z", "content": "Very well said.
", "rawMarkdown": "Very well said.", "votes": -3 }, { "id": 1084362, "postDate": "2020-11-20T01:01:26.823Z", "content": "well maid!
", "rawMarkdown": "well maid!" }, { "id": 596868, "postDate": "2019-08-11T12:31:01.283Z", "content": "Wow,, it's literally amazing
", "rawMarkdown": "Wow,, it's literally amazing" }, { "id": 260723, "postDate": "2017-12-20T19:16:27.517Z", "content": "Why not make the competition have no award to make it fair? You could instead give more virtual awards like more gold metals or some kind of special new \"diamond\" metal? Gamification works...
", "rawMarkdown": "Why not make the competition have no award to make it fair? You could instead give more virtual awards like more gold metals or some kind of special new \"diamond\" metal? Gamification works..." }, { "id": 256016, "postDate": "2017-12-11T00:03:27.783Z", "content": "The not host at all option is by far the better option. This is pure data greed.
", "rawMarkdown": "The not host at all option is by far the better option. This is pure data greed." }, { "id": 1029261, "postDate": "2020-09-27T15:48:25.307Z", "rawMarkdown": "", "isDeleted": true }, { "id": 201231, "postDate": "2017-07-10T18:33:36.117Z", "rawMarkdown": "", "votes": 2, "isDeleted": true } ], "comments": [ { "id": 200775, "author_name": "Guillermo Barbadillo", "author_url": "", "post_date": "2017-07-09T08:42:22.337000", "content": "Anthony, according to COMPETES ACT an american entity can receive the prize. So Kaggle could be that entity that receives the prize and later pays to non American people.
\n\nIt's as simple as that.
", "votes": 63, "replies": [ { "id": 200952, "author_name": "", "author_url": "", "post_date": "2017-07-10T01:47:36.793000", "content": "", "votes": 1, "replies": [] }, { "id": 205007, "author_name": "Anurag", "author_url": "", "post_date": "2017-07-20T12:29:04.537000", "content": "Why such a good suggestion by ironbar is being ignored.
", "votes": 3, "replies": [] }, { "id": 208725, "author_name": "bitbot", "author_url": "", "post_date": "2017-07-30T19:17:28.133000", "content": "are you certain that Kaggle is an American entity?
", "votes": 1, "replies": [] }, { "id": 209853, "author_name": "steven", "author_url": "", "post_date": "2017-08-03T15:55:58.077000", "content": "If it were that simple, the COMPETES act would have been formulated differently just to make sure that it weren't that easy to circumvent..
\n// Yep this statement loops on itself :P
For Kaggle to do that, Kaggle will need employment permission for each non-USA participant for this challenge.
", "votes": 0, "replies": [] }, { "id": 229036, "author_name": "Lalit", "author_url": "", "post_date": "2017-10-08T15:34:33.750000", "content": "If a solution by a non-USA participant is better than USA participants, Kaggle may purchase the solution from that participant, \"bottle\" it, and sell it. A win-win situation!
", "votes": 0, "replies": [] } ] }, { "id": 200691, "author_name": "Miha Skalic", "author_url": "", "post_date": "2017-07-08T22:56:56.540000", "content": "Anthony,\nYour (only) argument is that the “availability of datasets” to push forward ML. Yet, competition rules say:
\n\n\n\n\nThese images are for competition participants only. You should not share the images publicly, plan to use them in future research, or blog about them. You are also asked to destroy the data once the competition is over.
\n
Care to elaborate?
\n\nI believe you are bluffing here – I believe the sole reason is Kaggle's financial gain.
", "votes": 53, "replies": [ { "id": 204099, "author_name": "Timothy Rimnac", "author_url": "", "post_date": "2017-07-17T15:39:22.493000", "content": "Wow that's something that I didn't notice, I'm glad you pointed it out.
", "votes": 1, "replies": [] }, { "id": 206681, "author_name": "RobertoR", "author_url": "", "post_date": "2017-07-25T02:46:45.790000", "content": "", "votes": 1, "replies": [] } ] }, { "id": 201953, "author_name": "Cut Onion", "author_url": "", "post_date": "2017-07-11T16:16:15.973000", "content": "\n\n\nIf we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning.
\n
If so, make this dataset really available for us to learn. Currently it is available at terms like \"give us everything as community, we give nothing (to most of you)\". Allow unrestricted usage of data and we done (maybe, I couldn't talk for everybody here).
", "votes": 13, "replies": [] }, { "id": 202382, "author_name": "EduardoLi", "author_url": "", "post_date": "2017-07-12T12:02:25.300000", "content": "Anthony, a solution adopted by some sport competitions could solve the misunderstanding. To clarify, let us take the example of a city hosting an international marathon and the mayor wants to motivate local runners to participate. The fastest runner wins the general prize from the global sponsors and the fastest runner from that city wins an additional prize.\nThis strategy maintains the global competition feature and allows governments, institutions and companies to provide a local stimulus following the specific criteria they want and respecting legal restrictions.\nIn this competition, the best model(s) could receive a prize from Kaggle and the best competitors who meet the entity criteria (America COMPETES Act of 2010) are qualified for an additional prize paid by U.S. Transportation Security Administration (TSA).\nBest Regards,\nEduardo.
", "votes": 7, "replies": [ { "id": 202531, "author_name": "eagle4", "author_url": "", "post_date": "2017-07-12T18:08:59.003000", "content": "good points!\nMore extreme: no cash prizes, everybody can participate, everybody receives Kaggle points.
", "votes": 0, "replies": [] } ] }, { "id": 220913, "author_name": "Joshua Wonser", "author_url": "", "post_date": "2017-09-13T13:29:33.637000", "content": "Could you make it if a non-American wins they can donate their winnings to a charity?
", "votes": 4, "replies": [] }, { "id": 203211, "author_name": "MuratAykut", "author_url": "", "post_date": "2017-07-14T10:46:46.453000", "content": "Can we use this dataset for academic purposes in the future?
", "votes": 4, "replies": [] }, { "id": 200597, "author_name": "عثمان", "author_url": "", "post_date": "2017-07-08T16:38:20.100000", "content": "All the controversy behind this competition / dataset / DHS data provider reminds of a Netflix series titled 'The Machine'. Spoiler alert - throughout the series, the creator of this over-arching, all-seeing AI battles with the knowledge that his creation explicitly is being used to kill people without due course of law (and in some instances, it even directly kills people itself). He rightfully understood that if he didn't do it, someone else more sinister would. His desire in building it was to use technology to stop other people from doing bad things... and for the most part, was very successful in doing this---depending on who got to define who the 'bad' people were, of course. Throughout the series, we sometimes learned that the people we felt were bad actually weren't, and it was in fact the good guys who were very bad, or vice versa.
\n\nAnyhow, I don't really have much to add to this discussion except my belief that as people... as data hackers, explorers, engineers, scientists, or whatever, it really is our moral responsibility to see through to the end of our actions. By that I mean, sure, it might be interesting data and in interesting problem; but just looking one step ahead at the immediate goal of wining a contest, acquiring some fame, coming into some money, etc. when the technology we're working with is exponential in nature, and the organizations we develop the technology for are national, and the implications of our work undoubtedly will effect hundreds of millions of people, kinda is deserving of at least an introspective discussion, if not a public one. When added on top of that is the culture conflict with the contest host, iono... =\\
", "votes": 3, "replies": [] }, { "id": 203555, "author_name": "saquib khan", "author_url": "", "post_date": "2017-07-15T13:40:30.110000", "content": "One workaround for non-US people can be to partner with a US person whom you can trust, let that person get the prize money and later you can share among yourselves.
", "votes": 1, "replies": [ { "id": 207778, "author_name": "DrPastu", "author_url": "", "post_date": "2017-07-27T12:22:16.817000", "content": "\"Members of the Kaggle community who are not United States Citizens or legal permanent residents at the time of entry are allowed to participate in the Competition but are not eligible to win prizes. If a team has one or more members who are not prize eligible, then the entire team is not prize eligible.\"
", "votes": 1, "replies": [] }, { "id": 244450, "author_name": "Florens de Wit", "author_url": "", "post_date": "2017-11-16T10:30:01.777000", "content": "I think partnering could be done implicitly; you work together but don't form an official team on Kaggle.\nIt's like forming a team with your family or friends but use one personal profile to post solutions. I'm not sure if that would be a violation of rules though.
", "votes": 0, "replies": [] } ] }, { "id": 207781, "author_name": "DrPastu", "author_url": "", "post_date": "2017-07-27T12:31:39.717000", "content": "\"DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.\" Perhaps you should ask Alphabet Inc legal team to look into it, maybe they will have more luck.
", "votes": 0, "replies": [] }, { "id": 657927, "author_name": "", "author_url": "", "post_date": "2019-10-25T14:17:32.460000", "content": "well done
", "votes": -1, "replies": [] }, { "id": 209725, "author_name": "GillesDqn", "author_url": "", "post_date": "2017-08-03T07:05:06.380000", "content": "It is a well known fact that Google \"lacks\" money and cannot subsidize a prize in a non discriminatory way like \"should a non US citizen member of our community win the competition, we'd make sure he/she'd still happy to win\"... just saying, on the other hand, it might be a bit more complicated than that.
\n\nThis being said, discriminatory compromises even for the \"sake of a dataset\" have only taught one thing: it is ok to discriminate.
", "votes": -1, "replies": [] }, { "id": 252567, "author_name": "", "author_url": "", "post_date": "2017-12-03T08:48:10.707000", "content": "I have informations from the officers that are monitoring this individual Anthony Goldbloom for creating a gamification game called Kaglle , that nearly 99% of the competitions are viciated by information leks and the solutions trucate , the winners are well know even before the competition is hosted .
", "votes": -6, "replies": [] }, { "id": 204188, "author_name": "djona", "author_url": "", "post_date": "2017-07-17T20:53:22.100000", "content": "Very well said.
", "votes": -3, "replies": [] }, { "id": 1084362, "author_name": "Woojin Son", "author_url": "", "post_date": "2020-11-20T01:01:26.823000", "content": "well maid!
", "votes": 0, "replies": [] }, { "id": 596868, "author_name": "Edwin Yoon", "author_url": "", "post_date": "2019-08-11T12:31:01.283000", "content": "Wow,, it's literally amazing
", "votes": 0, "replies": [] }, { "id": 260723, "author_name": "", "author_url": "", "post_date": "2017-12-20T19:16:27.517000", "content": "Why not make the competition have no award to make it fair? You could instead give more virtual awards like more gold metals or some kind of special new \"diamond\" metal? Gamification works...
", "votes": 0, "replies": [] }, { "id": 256016, "author_name": "poke53280", "author_url": "", "post_date": "2017-12-11T00:03:27.783000", "content": "The not host at all option is by far the better option. This is pure data greed.
", "votes": 0, "replies": [] }, { "id": 1029261, "author_name": "", "author_url": "", "post_date": "2020-09-27T15:48:25.307000", "content": "", "votes": 0, "replies": [] }, { "id": 201231, "author_name": "", "author_url": "", "post_date": "2017-07-10T18:33:36.117000", "content": "", "votes": 2, "replies": [] } ], "raw_markdown_by_id": { "200775": "Anthony, according to COMPETES ACT an american entity can receive the prize. So Kaggle could be that entity that receives the prize and later pays to non American people. \n\nIt's as simple as that.", "200691": "Anthony,\nYour (only) argument is that the “availability of datasets” to push forward ML. Yet, competition rules say:\n\n>These images are for competition participants only. You should not share the images publicly, plan to use them in future research, or blog about them. You are also asked to destroy the data once the competition is over.\n\nCare to elaborate? \n\nI believe you are bluffing here – I believe the sole reason is Kaggle's financial gain. \n", "201953": "> If we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning. \n\nIf so, make this dataset really available for us to learn. Currently it is available at terms like \"give us everything as community, we give nothing (to most of you)\". Allow unrestricted usage of data and we done (maybe, I couldn't talk for everybody here).", "200588": "I understand and agree with frustrations around non-US citizens/permanent residents being ineligible for prize money. None of Kaggle, DHS or the community wants this restriction.\n\nThe reason for the restriction is the America COMPETES Act. DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.\n\nFrom Kaggle's perspective, we had a choice: host this competition with a geo restriction or not host at all. This was a tough choice. Hosting the competition meant running a two-speed competition, where some are eligible for prize money while others are not. Doing this violates our desire for meritocracy, where we offer equal opportunities to every community member regardless of education, background and (of course) country.\n\nIf we don't host, then we don't expose our community to one of the most interesting and valuable datasets that's ever been hosted on Kaggle. As we've seen over the past five years with the rise of deep neural networks, the availability of datasets allows us to push forward to science of machine learning. Not exposing this dataset to our community meant depriving our community, and the machine learning world more generally, of a chance to push forward the science with a novel and challenging dataset.\n\nWe decided that it was better to make the dataset available. Many in this forum have said this was a mistake. This was a challenging issue; I hope this at least gives you a window into our thinking.\n\nAnthony", "202382": "Anthony, a solution adopted by some sport competitions could solve the misunderstanding. To clarify, let us take the example of a city hosting an international marathon and the mayor wants to motivate local runners to participate. The fastest runner wins the general prize from the global sponsors and the fastest runner from that city wins an additional prize.\nThis strategy maintains the global competition feature and allows governments, institutions and companies to provide a local stimulus following the specific criteria they want and respecting legal restrictions.\nIn this competition, the best model(s) could receive a prize from Kaggle and the best competitors who meet the entity criteria (America COMPETES Act of 2010) are qualified for an additional prize paid by U.S. Transportation Security Administration (TSA).\nBest Regards,\nEduardo.", "220913": "Could you make it if a non-American wins they can donate their winnings to a charity?", "203211": "Can we use this dataset for academic purposes in the future?", "200597": "All the controversy behind this competition / dataset / DHS data provider reminds of a Netflix series titled 'The Machine'. Spoiler alert - throughout the series, the creator of this over-arching, all-seeing AI battles with the knowledge that his creation explicitly is being used to kill people without due course of law (and in some instances, it even directly kills people itself). He rightfully understood that if he didn't do it, someone else more sinister would. His desire in building it was to use technology to stop other people from doing bad things... and for the most part, was very successful in doing this---depending on who got to define who the 'bad' people were, of course. Throughout the series, we sometimes learned that the people we felt were bad actually weren't, and it was in fact the good guys who were *very* bad, or vice versa.\n\nAnyhow, I don't really have much to add to this discussion except my belief that as people... as data hackers, explorers, engineers, scientists, or whatever, it really is our moral responsibility to see through to the end of our actions. By that I mean, sure, it might be interesting data and in interesting problem; but just looking one step ahead at the immediate goal of wining a contest, acquiring some fame, coming into some money, etc. when the technology we're working with is exponential in nature, and the organizations we develop the technology for are national, and the implications of our work undoubtedly will effect hundreds of millions of people, kinda is deserving of at least an introspective discussion, if not a public one. When added on top of that is the culture conflict with the contest host, iono... =\\", "203555": "One workaround for non-US people can be to partner with a US person whom you can trust, let that person get the prize money and later you can share among yourselves.", "207781": "\"DHS's legal team worked hard to find a way to allow international participation at all (albeit without prize money) while not violating the act.\" Perhaps you should ask Alphabet Inc legal team to look into it, maybe they will have more luck.", "657927": "well done", "209725": "It is a well known fact that Google \"lacks\" money and cannot subsidize a prize in a non discriminatory way like \"should a non US citizen member of our community win the competition, we'd make sure he/she'd still happy to win\"... just saying, on the other hand, it might be a bit more complicated than that.\n\nThis being said, discriminatory compromises even for the \"sake of a dataset\" have only taught one thing: it is ok to discriminate.", "252567": "I have informations from the officers that are monitoring this individual Anthony Goldbloom for creating a gamification game called Kaglle , that nearly 99% of the competitions are viciated by information leks and the solutions trucate , the winners are well know even before the competition is hosted .", "204188": "Very well said.", "1084362": "well maid!", "596868": "Wow,, it's literally amazing", "260723": "Why not make the competition have no award to make it fair? You could instead give more virtual awards like more gold metals or some kind of special new \"diamond\" metal? Gamification works...", "256016": "The not host at all option is by far the better option. This is pure data greed.", "1029261": "", "201231": "" } }