5 ms·

what is the privacy loss outcome that you are suggesting has happened as a result of photos being disclosed to algorithms?

bluntfang 6y ago
use your imagination. Sheesh this is such a boring line of questioning that borders on disingenuous and obtuse for a social news site that caters to people in AdTech.
- dang 6y ago
  You've unfortunately been breaking the site guidelines a lot, by using HN for political and ideological battle, flaming other users, and so on. We've had to warn you about this many times: https://news.ycombinator.com/item?id=23452134 https://news.ycombinator.com/item?id=23452134 https://news.ycombinator.com/item?id=21480233 https://news.ycombinator.com/item?id=21480233 https://news.ycombinator.com/item?id=21480058 https://news.ycombinator.com/item?id=21480058 https://news.ycombinator.com/item?id=20717150 https://news.ycombinator.com/item?id=20717150 Would you please fix this? The intended use of the site is curious, thoughtful conversation. I'm not going to ban you right now because (not counting the last week) you've posted good, on-topic comments too. But if you continue along the same lines we're going to have to. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
  - bluntfang 6y ago
    That's interesting that you seem to think it's my problem. I also find it disingenuous that you are linking to my previous comments publically. Are you trying to shame me? Maybe you need to disallow political discussions period? Have you considered that the people I am replying to are also fighting a political battle that you are biased for? All in all, yes I believe certain things passionately. Like racism and sexist exist in America, specifically in software and corporate America. I also believe that this bias shows on this community because of the demographic or people who browse and comment here.
    - dang 6y ago
      I'm not trying to shame you; sorry if it had that effect. I post links like that when I imagine it might be helpful to users who are trying to decide (as users always try to decide) if the moderator is being excessive or not. Banning political discussions is not possible for a bunch of reasons, but the most important one is that it would be bad for curiosity, which is the thing we try to optimize HN for (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sort=byDate&type=comment&query=curiosity%20optimiz%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...). There are too many intellectually interesting topics with political overlap—and ultimately, some political overlap on virtually every topic. At the same time, not moderating political discussions is also not an option because that would also be bad for curiosity. So we're in a bit of a tradeoff with constraints on either side and no easy solution. I've written about this a lot in the past: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&sort=byDate&type=comment&query=political%20overlap%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so.... If you or anyone reads those explanations and still has a question that I haven't answered yet, I'd like to know what it is. As for bias, everyone always feels that the mods are biased against their position when we moderate this way (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&sort=byDate&type=comment&query=%22always%20feels%22%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...). I'd never claim we're unbiased—who could?—but the bar for reaching that conclusion needs to be a little higher. I can show you a long, long list of HN commenters complaining bitterly that HN is biased in favor of your positions (i.e. the politics/ideology you happen to favor). Every passionate user feels that the bias runs against their side. I'd be cautious about assuming things about the "demographic or people who browse and comment here". We don't have that data—no one does—and HN is basically a gigantic projection screen or Rohrschach diagram out of which people draw their own pre-existing positions. (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=by%3Adang%20retina&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...) I know this because everyone does this, and the universal property their statements have in common is not their conclusions about HN (those are wildly contradictory) but the fact that they always see HN—both mods and community—as opposed to their own view. (One explanation for this is https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sort=byDate&type=comment&query=notice%20dislike%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...)
baq 6y ago
Network effects, unintended side effects, operator incompetence, risk of information leak... GDPR should have me covered, but I won’t be able to remove myself from deep learning models they already have.
- asdfasgasdgasdg 6y ago
  That is not true. Google retrains its models periodically. Removed data is not used for training models. Also I'm not sure that private user photos can even be used for training cross user models because I don't see how these photos could possibly be disclosed to raters, so there would be no ground truth.

bluntfang 6y ago

use your imagination. Sheesh this is such a boring line of questioning that borders on disingenuous and obtuse for a social news site that caters to people in AdTech.

dang 6y ago
You've unfortunately been breaking the site guidelines a lot, by using HN for political and ideological battle, flaming other users, and so on. We've had to warn you about this many times: https://news.ycombinator.com/item?id=23452134 https://news.ycombinator.com/item?id=23452134 https://news.ycombinator.com/item?id=21480233 https://news.ycombinator.com/item?id=21480233 https://news.ycombinator.com/item?id=21480058 https://news.ycombinator.com/item?id=21480058 https://news.ycombinator.com/item?id=20717150 https://news.ycombinator.com/item?id=20717150 Would you please fix this? The intended use of the site is curious, thoughtful conversation. I'm not going to ban you right now because (not counting the last week) you've posted good, on-topic comments too. But if you continue along the same lines we're going to have to. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- bluntfang 6y ago
  That's interesting that you seem to think it's my problem. I also find it disingenuous that you are linking to my previous comments publically. Are you trying to shame me? Maybe you need to disallow political discussions period? Have you considered that the people I am replying to are also fighting a political battle that you are biased for? All in all, yes I believe certain things passionately. Like racism and sexist exist in America, specifically in software and corporate America. I also believe that this bias shows on this community because of the demographic or people who browse and comment here.
  - dang 6y ago
    I'm not trying to shame you; sorry if it had that effect. I post links like that when I imagine it might be helpful to users who are trying to decide (as users always try to decide) if the moderator is being excessive or not. Banning political discussions is not possible for a bunch of reasons, but the most important one is that it would be bad for curiosity, which is the thing we try to optimize HN for (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sort=byDate&type=comment&query=curiosity%20optimiz%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...). There are too many intellectually interesting topics with political overlap—and ultimately, some political overlap on virtually every topic. At the same time, not moderating political discussions is also not an option because that would also be bad for curiosity. So we're in a bit of a tradeoff with constraints on either side and no easy solution. I've written about this a lot in the past: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&sort=byDate&type=comment&query=political%20overlap%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so.... If you or anyone reads those explanations and still has a question that I haven't answered yet, I'd like to know what it is. As for bias, everyone always feels that the mods are biased against their position when we moderate this way (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&sort=byDate&type=comment&query=%22always%20feels%22%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...). I'd never claim we're unbiased—who could?—but the bar for reaching that conclusion needs to be a little higher. I can show you a long, long list of HN commenters complaining bitterly that HN is biased in favor of your positions (i.e. the politics/ideology you happen to favor). Every passionate user feels that the bias runs against their side. I'd be cautious about assuming things about the "demographic or people who browse and comment here". We don't have that data—no one does—and HN is basically a gigantic projection screen or Rohrschach diagram out of which people draw their own pre-existing positions. (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=by%3Adang%20retina&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...) I know this because everyone does this, and the universal property their statements have in common is not their conclusions about HN (those are wildly contradictory) but the fact that they always see HN—both mods and community—as opposed to their own view. (One explanation for this is https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sort=byDate&type=comment&query=notice%20dislike%20by:dang https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...)

baq 6y ago

Network effects, unintended side effects, operator incompetence, risk of information leak... GDPR should have me covered, but I won’t be able to remove myself from deep learning models they already have.

asdfasgasdgasdg 6y ago
That is not true. Google retrains its models periodically. Removed data is not used for training models. Also I'm not sure that private user photos can even be used for training cross user models because I don't see how these photos could possibly be disclosed to raters, so there would be no ground truth.