3 ms·

Have you thought about how you're going to "dickproof" this? What's the plan for when trolls descend and start using it to anonymously host hate lies or abuse?

z3t4 6y ago
Do things that dont scale. Most problems can be solved. The most difficult one is making ppl love your service. eg. gaining users.
- bigiain 6y ago
  Parler has just demonstrated this problem specifically cannot always be (easily) solved later. By the time a noticeable subset of your users are doing things your investors/advertisers/hosting providers/app stores/mother would consider "deplorable", it's _way_ too late to be thinking "Yeah, maybe we should have built in some protection for ordinary users against griefers/trolls/russian and chinese misinformation campaigns." I'm not saying they need to do it right now, or that it's a more important thing than gaining users/traction right now, but it should without doubt be a thing that's very near the top of their list of "things we have considered and have vague plans to implement, that need to watch carefully for, and to have not painted ourselves into corners where our ideas about what we'll need to do to stamp it out no longer work any more..."
  - saimiam 6y ago
    I feel this is a bit of a stretch. Parler was positioned as a free speech fundamentalist so it never stood a chance when it came time to moderate content to stay in the good books of its vendors. Otoh, seems to me that Tusk is "micro social" - a term I just invented - to socialize a page/concept amongst a smallish group of people. Unless that link leaks, how will the online trolls even find a page to tusk? And even if does, adding an auth layer on top of page tied to your identity should be relatively easy. FD - I am working on a "public" micro-social offering as well (info in profile) so how Tusk addresses this is of interest to me.
    - canadianfella 6y ago
      What is a Parler vendor?
  - skinkestek 6y ago
    > "Yeah, maybe we should have built in some protection for ordinary users against griefers/trolls/russian and chinese misinformation campaigns." Parler was killed not because of that but because they didn't manage to protect themselves against the incumbents. Parler might have been awful, IDK, but both Facebook and Twitter has been really awful without being taken down. In fact the only place I've seen actual nazi content, e.g. "certain groups of people should be exterminated, what a shame Hitler died so early"-kind-of-comments without looking for it has been in Facebook comments.
    - ycombigator 6y ago
      Don't challenge the narrative, you'll get cancelled.
  - ycombigator 6y ago
    No more so than Twitter.
- afarrell 6y ago
  Okay, what is the "thing that does not scale" which they are starting out with? There's a difference among: A) We are having someone in the company moderate this manually and gain insight into how we could do so scalably. B) We are outsourcing moderation and therefore not gaining insight into how to reduce the burden of that cost centre. C) We aren't worrying about that until later. D) We are building a fully-automated solution from the beginning.
brigandish 6y ago
Maybe I'm missing something, but wouldn't the ephemeral nature would make that difficult? (or more difficult than on something like Twitter)
- bigiain 6y ago
  3 days is like 9 or 10 "news cycles" these days. In a world of viral social media content, I wouldn't consider this to be particularly ephemeral.
n_david 6y ago
I talked a bit about this in my comment above in response to sanmak's comment, but the short answer is I don't really know the best way to prevent abuse on the platform and I'm hoping that the short-lived feedback posts and the sort of pseudo-anonymous domains at least helps prevent it. Like saimiam mentioned, the domains aren't exactly meant to be shared on the internet for the entire world to pitch in on the topic. It's more meant for the domain creator to share with people whom they would like to receive feedback from, people whom they trust. Honestly, abuse is really the absolute last thing that I want for this platform to give a voice to. I'm still thinking of ways to best prevent this, but if it really devolves into a platform for people to spread hate, I will just shut it down.
- bigiain 6y ago
  I strongly suggest you implement some sort of monitoring on Twitter/Facebook/Reddit/TikTok et al. for your pseudo-anonymous domains. Either pro-active monitoring of "the firehouse" if available/affordable, or at least reactive monitoring of inbound referrers and unusually large traffic volumes to individual subdomains. Please don't "just shut it down", fight it. At least try. We need more "good things" on the internet that don 't/can't get fucked up by people being assholes. If your intention is "the domains aren't exactly meant to be shared on the internet for the entire world to pitch in ", maybe enforce that right up front? Ratelimit requests perhaps, so the first hundred or two requests are fast, but that it obviously and quickly slows down and becomes unusable above a certain size? Keep the free/anonymous service to a size that intentionally breaks well before it is involved in any (potentially problematic) viral social media stuff. That might be a future path to monetisation even? Tweak your thresholds to it's fast and free for the first half a dozen or so people you share it with, but slows dramatically if you try and send a lot more people the link. Then you could offer paid plans with increased limits, you can then use the existence of a paying identity to avoid/block people who misuse the system? Even a nominal ~$5/year payment would dissuade griefers if you made it clear you'd supply payment identities to proper legal requests. And if your service proves wildly popular, selling "enterprise" account to to brands who want to use in in their viral content" attempts could perhaps not make you a unicorn, but could become a nice "lifestyle business". That'd spoil your HN cred, but "new Corolla every year" money from a side hustle is perhaps sometimes better that "swinging for the fences for dreams of Lambo money" gambles...

z3t4 6y ago

Do things that dont scale. Most problems can be solved. The most difficult one is making ppl love your service. eg. gaining users.

bigiain 6y ago
Parler has just demonstrated this problem specifically cannot always be (easily) solved later. By the time a noticeable subset of your users are doing things your investors/advertisers/hosting providers/app stores/mother would consider "deplorable", it's _way_ too late to be thinking "Yeah, maybe we should have built in some protection for ordinary users against griefers/trolls/russian and chinese misinformation campaigns." I'm not saying they need to do it right now, or that it's a more important thing than gaining users/traction right now, but it should without doubt be a thing that's very near the top of their list of "things we have considered and have vague plans to implement, that need to watch carefully for, and to have not painted ourselves into corners where our ideas about what we'll need to do to stamp it out no longer work any more..."
- saimiam 6y ago
  I feel this is a bit of a stretch. Parler was positioned as a free speech fundamentalist so it never stood a chance when it came time to moderate content to stay in the good books of its vendors. Otoh, seems to me that Tusk is "micro social" - a term I just invented - to socialize a page/concept amongst a smallish group of people. Unless that link leaks, how will the online trolls even find a page to tusk? And even if does, adding an auth layer on top of page tied to your identity should be relatively easy. FD - I am working on a "public" micro-social offering as well (info in profile) so how Tusk addresses this is of interest to me.
  - canadianfella 6y ago
    What is a Parler vendor?
- skinkestek 6y ago
  > "Yeah, maybe we should have built in some protection for ordinary users against griefers/trolls/russian and chinese misinformation campaigns." Parler was killed not because of that but because they didn't manage to protect themselves against the incumbents. Parler might have been awful, IDK, but both Facebook and Twitter has been really awful without being taken down. In fact the only place I've seen actual nazi content, e.g. "certain groups of people should be exterminated, what a shame Hitler died so early"-kind-of-comments without looking for it has been in Facebook comments.
  - ycombigator 6y ago
    Don't challenge the narrative, you'll get cancelled.
- ycombigator 6y ago
  No more so than Twitter.
afarrell 6y ago
Okay, what is the "thing that does not scale" which they are starting out with? There's a difference among: A) We are having someone in the company moderate this manually and gain insight into how we could do so scalably. B) We are outsourcing moderation and therefore not gaining insight into how to reduce the burden of that cost centre. C) We aren't worrying about that until later. D) We are building a fully-automated solution from the beginning.

brigandish 6y ago

Maybe I'm missing something, but wouldn't the ephemeral nature would make that difficult? (or more difficult than on something like Twitter)

bigiain 6y ago
3 days is like 9 or 10 "news cycles" these days. In a world of viral social media content, I wouldn't consider this to be particularly ephemeral.

n_david 6y ago

I talked a bit about this in my comment above in response to sanmak's comment, but the short answer is I don't really know the best way to prevent abuse on the platform and I'm hoping that the short-lived feedback posts and the sort of pseudo-anonymous domains at least helps prevent it. Like saimiam mentioned, the domains aren't exactly meant to be shared on the internet for the entire world to pitch in on the topic. It's more meant for the domain creator to share with people whom they would like to receive feedback from, people whom they trust. Honestly, abuse is really the absolute last thing that I want for this platform to give a voice to. I'm still thinking of ways to best prevent this, but if it really devolves into a platform for people to spread hate, I will just shut it down.

bigiain 6y ago
I strongly suggest you implement some sort of monitoring on Twitter/Facebook/Reddit/TikTok et al. for your pseudo-anonymous domains. Either pro-active monitoring of "the firehouse" if available/affordable, or at least reactive monitoring of inbound referrers and unusually large traffic volumes to individual subdomains. Please don't "just shut it down", fight it. At least try. We need more "good things" on the internet that don 't/can't get fucked up by people being assholes. If your intention is "the domains aren't exactly meant to be shared on the internet for the entire world to pitch in ", maybe enforce that right up front? Ratelimit requests perhaps, so the first hundred or two requests are fast, but that it obviously and quickly slows down and becomes unusable above a certain size? Keep the free/anonymous service to a size that intentionally breaks well before it is involved in any (potentially problematic) viral social media stuff. That might be a future path to monetisation even? Tweak your thresholds to it's fast and free for the first half a dozen or so people you share it with, but slows dramatically if you try and send a lot more people the link. Then you could offer paid plans with increased limits, you can then use the existence of a paying identity to avoid/block people who misuse the system? Even a nominal ~$5/year payment would dissuade griefers if you made it clear you'd supply payment identities to proper legal requests. And if your service proves wildly popular, selling "enterprise" account to to brands who want to use in in their viral content" attempts could perhaps not make you a unicorn, but could become a nice "lifestyle business". That'd spoil your HN cred, but "new Corolla every year" money from a side hustle is perhaps sometimes better that "swinging for the fences for dreams of Lambo money" gambles...