Dan Romero on Warpcast

Dan Romero pfp

The spam filtering dilemma Sharing how we think about it so people can better understand: 1. Most people will only use social networks if they are fun 2. Casting something and getting dozens of spammy replies is not fun 3. If you get enough spammy replies, you'll just stop using the app 4. So, aggressively filtering for spam is necessary to keep your existing users happy 5. However, this means some good new users will hit spam filters. Understatement: this is a bad user experience for those people. It would be great to not have this happen. 6. But given the choice of who to make unhappy — an existing user who has invested a ton of time and energy into the network or a new user that is most likely to churn (top tier social networks lose 50% of users!), then you have to optimize for your existing users. 7. So does that mean you don’t care about new users? No! It’s existential to have new users be able to join Farcaster and find people to connect with — without being labeled as spam. 1/2

32 replies

87 recasts

437 reactions

Dan Romero pfp

8. So here’s what we are doing to improve this so new good users aren’t mislabeled as spam: a. Continually refine the spam model by identifying incorrectly labeled people (that feeds back into the model and makes it better); good existing users on the network reporting incorrect labels is really valuable here b. Warnings to users before they are labeled spam so they have a chance to adjust their behavior c. Increase the number of signals our model uses to help filter out bots vs. people d. Working a new version of channels that give owners / moderators complete control of who appears in their channel (so a Warpcast spam label won’t affect casts in the channel) 2/2

16 replies

31 recasts

223 reactions

Daisy✨ pfp

On 8a So if we notice real users being labeled wrongfully we should share right? I’ve noticed about 2 genuine users who are marked Is it okay if I share their usernames under this cast?

1 reply

0 recast

4 reactions

Sophia Indrajaal pfp

Sophia Indrajaal

@sophia-indrajaal

With the Vietnamese Mask Afficionado Bot Society, I think it is meant to game the system. They all reply to each other, which with my limited understanding seems to imply there is enough Moxie or whatnot generated to pay for new accounts. The goal is spammy replies in that case, which kind of creates a looped subsection of Farcaster that isn't even meant for much interaction with other accounts except maybe on occasion to build up their stats. Not sure if this adds much to the conversation, but leaving it here just in case

1 reply

0 recast

2 reactions

Tim pfp

I don’t get it Sometimes when I open the see more replies a bot will be way off and it makes me laugh To each their own though

1 reply

0 recast

1 reaction

Mikado pfp

A report as not spam button could help you refine the algorithm. This could also weight the reports by the reputation of the person who reported.

1 reply

0 recast

1 reaction

Alemac pfp

sounds very good! But what happens to people who have already been tagged as spam?

1 reply

0 recast

0 reaction

👑eiteen go🎅🎄<3 pfp

👑eiteen go🎅🎄<3

8. ⇒If someone's message has been mistakenly classified as spam, where should I officially contact?

1 reply

0 recast

0 reaction

raulonastool.eth 🎩 🏰 pfp

raulonastool.eth 🎩 🏰

Respect this approach. fair for all. Would be nice to also get some transparency on the "spam filtering algorithm" (if it exists). Will there be Docs shared somewhere?

1 reply

0 recast

0 reaction

binocularsXL🚀 pfp

binocularsXL🚀

Concerning 8d Will the channels plan be following what Jason posted a few days back?

1 reply

0 recast

0 reaction

Stephan pfp

glad you're prioritizing the problem!

0 reply

0 recast

12 reactions

Maretus pfp

Good move Dan. Appreciate that you’re thinking about this. If the vouch system uses a system like openrank to determine who can vouch, I think it could be effective. Only let people who have something to actually lose do it. That will create a system that self polices better. Penalty for vouching for a spam user = no more ability to vouch for X amount of time.

0 reply

0 recast

1 reaction

DenizenJuan.⌐◨-◨ テ pfp

DenizenJuan.⌐◨-◨ テ

How about proof of humanity? Gitcoin Passport, BrightID, or the other project that scans the iris that I don’t like, etc. More proof = more boost Less proof = less visible

0 reply

0 recast

2 reactions

Push 🎩🤌🏻 pfp

Push 🎩🤌🏻

This is good💆🏻‍♂️ Definitely prioritizing the ogs a bit, but I’m on the winning side he so my voice doesn’t really count. Still we need to find a better way maybe to educate a new user. Maybe something during the onboarding to preventively let them understand how it works

0 reply

0 recast

0 reaction

Breck Yunits pfp

There's a far smarter approach.

0 reply

0 recast

0 reaction

law pfp

Dan about this whole new user being labeled thing Mike Rainbow helped me twice to bring your attention to my friends who i onboarded and ended up in spam for like no reason But nothing was done. I know you have a lot you work on and i hope you see this Again

0 reply

0 recast

1 reaction

CheddarQueso pfp

Most of my spam is in the form of useless frames, token pushing and engagement farming. I post genuine and get less views here than on any other platform. Not sure how to connect. 🫠

0 reply

0 recast

0 reaction

aintnothinxart pfp

@aintnothinxart

there's been a huge wave of AI responses to all my posts over the last few weeks.

0 reply

0 recast

0 reaction

jrcart.base.eth pfp

jrcart.base.eth

I have been getting a lot of spam replies lately. So we definitely need something. Blows my mind people will pay to sign up a bot here!

0 reply

0 recast

0 reaction

Cristian 🎩 pfp

@cristianisbased.eth

Warnings are a step in the right direction But I think there should be a way to vouch for someone labeled as spam to be given another chance as well There will always be cracks in any system and good apples that are mistaken for bad...

0 reply

0 recast

0 reaction