How new teams can build community around the product on PH?
I’ve been a part of PH for 2+ years now, and I really love it, met real friends like @abdalyousef and @michalhajtas, and I really learn from @busmark_w_nika , @abbas143official . My team (Scade) have launched twice, and we've been active supporting other launches, giving feedback, sharing advice in discussion, connecting with other makers, and speaking about our own product. This helped us become visible on the platform, get feedback before launch and be known to the community on launch day. It lead us to a Product of the day, twice, and even a nomination for Kitty awards.
Now, I’m helping a new product with PH launch. Guys are new to PH and they genuinely got involved. They filled out their profiles with real names, photos, bios, and links. They started testing products and giving feedback to makers and asking question, reading discussion, and sharing their thoughts.
But no one can see them. They created accounts about two weeks ago, comment by themselves without AI, but their comments seem useless for PH or what? Product Hunt support keeps pointing to the guidelines, but everything in those rules is followed. Comments are flagged as AI-generated, which is unfair (funny cause this is the team of AI detection experts and the product we aim to launch is on AI detection).
Personally I think that this creates a gap between new makers and old ones, who have their second or third launch.
I don’t understand how these moderation algorithms work, and why real people with filled profiles are hidden. I believe it's almost impossible to be seen and make a community without supporting others, but nobody sees their support.
I want to ask the PH team how do you flag ai content? as an AI detection team, we know what is a false positive and that any ai detector can't be 100% sure in content origin.
And another question for new makers, how did you get noticed, and did you struggle with this? (hopefully mods will let this discussion live + you comments will be visible here)
Replies
Yeah. I wont say flagging AI content, rather flagging dumb content would be more prudent.
Scade.pro
@0x_ash11 Or something rude or spammy. I completely understand why we have these AI restrictions. I don't like bots in comments either, but there must be transparency, and real people shouldn't be flagged as AI.
@nastassia_k I completely agree.
IXORD
@0x_ash11 @nastassia_k But AI is being trained, and soon it will be difficult to distinguish it from a human. For example, even AI text scanners sometimes identify human-written text as AI-generated. However, the number of bots on PH has significantly decreased compared to before :)
@nastassia_k @ixord Igor, In my opinion; It is not that bad, till when the quality of content is great. I think a little different. I am afraid of automation bias.
The machine is as good as the operator. If you see that the quality of AI content is bad, then it is a important signal of declining IQ of an individual, and if I may, Society as whole.
Scade.pro
@ixord @0x_ash11 great point.
Scade.pro
@ixord @nastassia_k Yeah. I cant tell you how many points I have in my mind in For and Against arguments. If I start writing, I am afraid, It will become a reddit thread.
Scade.pro
Here's this great and almost 'invisible team' behind isFake who supported launches for two weeks, but you probably haven't seen them yet:
@olga_scry Olga, CPO
@frih Ivan, UX/UI designer who tested every product he commented on, but was flagged as AI, even though he didn't use it. We checked all of his comments using their AI detector and found no evidence of cheating..
@daria_ignatenko Daria, Data Scientist
@bandrw Andrei, Software Engineer
@maksim_avery Maksim, CMO and the only 'visible' guy in the team :)
minimalist phone: creating folders
Sorry for the late reply, I had a notification page under error for the past week, and now it is live, so I am returning to the comments and replying. TBH, I do not know how this works because there are a lot of changes here, and understand both parties.
Personally, I would apply some of the 3rd party tools for checking ID cards because I struggled a lot with bots. A 2 or 3 weeks ago reported over 100 profiles and sometimes had a problem differentiating whether it is AI or not.
@busmark_w_nika do you know any of such tools& I want to try them, just to check my team accounts
minimalist phone: creating folders
@olga_scry TBH, https://withpersona.com/ is used by LI, I recalled them among first.
Scade.pro
@busmark_w_nika thank you Nika, I see. Yes I totally understand the planform, nobody wants bots at their place. It's more about methods of detection
Scade.pro
Yes, I also hear about this problem (shadow ban) very often, and the responses from technical support were as follows: 1. your profile is poorly filled out (although other profiles with no information at all still receive comments)
2. your message is generated by AI (but no, it was written by hand)
3. your message is not informative or violates PH's policies (but no, the messages are relevant and do not violate anything).
4. the messages are being moderated, and the account is currently under moderation, which may take a week or two (this is more realistic).
It's a bit sad that users need to prove that they are not bots((
If there are any articles on this topic from ProductHunt, I'd be glad to read them (about algorithms and requirements for accounts and about shadow bans)
@maria_anosova Maria, thank you for this list. I think that my team accounts might be too "young" for posting comments, otherwise we did not violate any other rules. do we need to keep writing to support if the problem won't be solved?
I really want know as well, I once made a post asking what the best Product Hunt case you’ve seen is, but I don’t know why it got banned.
Scade.pro
@leah_l sorry to hear that Leah
I believe that thy only approve a small % of all discussions. Probably this question was already posted by someone else...