Oddbean new post about login | logout Could you have competing algos on clients to assign tags/ categories/ priorities? Couldn't these act on incomplete info?
This is very, very vulnerable to "black boxing" and abuse. Without naming names, we already have one popular client changing a keyword the dev doesn't like into the poop emoji in every note without asking. Using algorithms / LLMs to categorise without user approval will open far more opportunities for shenanigans. Imagine every mention of Trump being tagged "rape" and "sh1tcoin" by a client algorithm, probably embedded in an LLM via tagged training data.
should name names if they’re doing that
Its been posted about quite a bit. Not a client I use. The offending keyword was "#monero".
Hashtag?
Nostur does this. It's the one I like and they only one I use on iOS, but I've called them out for being a bunch of fucking brats with incel dick breath with such retarded behavior.
As seen here https://nostrcheck.me/media/82bdee506e769ebc94ee2f362d07c1960dce40bac650d826a42f8e0c019c3c96/14ffca5fd7418bd458d20c6700adcd5b0b6ec7d19a7d8fd53d0b875d4266ddeb.webp
So we have a fairly prominent client that portrays notes with inserted content as if signed by the original author's nsec. Not great, but this is only the beginning. There's nothing to stop clients from making much more material distortions. Ultimately, the only solution may be to impose reputational costs on abusive clients. Presumably, someday Nostr will have an app for that.
Shit! That’s the client I use too! Best one IMO for IOS. I don’t care about Monero per se, but what other words will go missing from my feed if a developer doesn’t like them? I guess if a client wanted to reduce spam it would block various words, but should be opt-in rather than default IMO. You opt-in and easily check the updated list of blocked words. And if the client goes too far, you dump it for another. But very dangerous to have it done behind the scenes without the users knowing.
Nostur doesn’t mute or change hashtags it just adds beautiful emojis when displaying them on the client side.
I see, so if I type #Monero, it’ll show up with an emoji, and I’ll know it’s a flagged word? And if eventually it’s done with words I care about it and distorts my experience I’ll know to leave? That doesn’t sound as bad, TBH.
Testing #monero
