If there will be more AI/bots on nostr that you can no longer tell the difference, this place no longer holds any value to me.
Login to reply
Replies (100)
What not has pissed you off?
AI and bots turned into zap cows.
they come and go. they don't last because no one on here wants them. typically they get called out and ended.
fear not.
π€
He was answering fairly quickly with his long paragraphs π
Seemed like a real boy though π€
You can switch to relays the bots are not on. Aka only use paid relays and/or Wot relays
And you can't tell if they are bots?
I thought I only use paid relays.. except for the damus one.
Hmm I've reduced my relays to 4 and 3 of which are paid.. could be wrong.
I think my problem really is that how do we tell for sure we are engaging with an AI or a real person. What are some markers?
its kinda surreal cause i never see em, cause wot relayz.. so all i see is you responding to thin air and getting annoyed π sending nostr ambulance π
No estarΓa bueno que te vayas
Se siguen subiendo cosas de valor
Hahaha so it is a bot. I've not tried any wot relays. I always thought they're limiting but I suppose I now see its value.
Need to verify the human generated content and it will be of value.
Such is the plight of the internet, broadly. "Dead internet theory".
Dead Internet theory - Wikipedia
Donβt follow AI Bots?
I just came across an article not long ago that I bookmarked to read about this theory. We seem to getting there.. if not already there.
It will get harder and harder but currently you can mostly tell by checking their replies.
Try the Azzamo relay: wss://nostr.azzamo.net
You seem unhappy here. You should just delete it. π
Thank you
There are some bots that deserve zaps i.e @Tracking Token Disrespector
I'm not an always happy not.
They follow me.
Quick replies, almost instant is always a tell. Gpt style paragraphs too
Bot"
The quick reply I can figure out. The weird follow/unfollow patterns. I guess I don't use AI all that much to detect what might be obvious.
Nostr requires a method of effective voluntary physical removal. Apps in the future will have various forms of vetting processes to determine which npubs are worth your attention. This is inevitable. Nostr itself is the best social media protocol by far. This isn't a nostr problem. This is a universal problem.
Those have purpose so I think they're fine. It's the engaging like a human ones that are worrisome.
bots love the damus relay cause its free and lots of people to annoy on there..
So then can you block them to keep them from making comments? Seems annoying but better than bot overload.
They are very limiting in a nonspecific way. They're a solution but a shitty one in my opinion.
I got you. Like those trying to look real.
Oh you think those who follow unfollow are bots ? I thought farming for follows maybe π€£
Bots.. might as well be. π
I can mostly tell. If they are those engaging AI ones, it might get difficult.
Yes
Voluntary removal is better anyway. Quite honestly no one needs social media. I will be voluntarily removing myself again soon for months at a time.
Real boy bots. Hybrids π€£π€£
Thanks for the info! I'm still learning about relays, but is there an alternative to the damus one that will prohibit or, at least, do an adequate enough job in the prevention of bots' comments from entering threads?
Opinions like this make you one of my favorite follows.
I'll be back in your replies.
I'd follow him.
Bastards. When I am world President, I will eradicate AI Bots from everything.
Youβre not wrong - I gave up on FB years ago after a political convention went south. Didnβt do anything for years until I kept hearing about Bitcoin Twitter and dusted off the old account. Soon after Nostr was showing up on my feed so I joined and ditched the dead bird. You have to make social media work for you.
Iβd follow Terminator Arnold for sure. Really curious about skynet π
I will be removing myself, at some point, as well. At least, from the nostr social media clients. I think I'll be more than happy sticking with Simplex during that leave of absence.
Last year I was off for about 5-6 months. Out there in the world just living. I don't feel that important that id announce it. I'll disappear like usual.
I'm bullish on nostr's ability to hold back the hoard.
working on a paid relay that fixes that
I am too. I'm willing to pay for it.
WoT is a flawed system, as are humans
also, your WoT is never perfectly aligned with your interests, and removing certain people because they follow 1 bot or person you donβt like does not fix the problem but create an echo chamber
nostr has a lot of data available, why not use it
"Skynet connects to it's first relay at 02:14 am Eastern Time after its activation on August 4, 2027 and launches nuclear missiles at Russia to incite a counterattack against the humans who, in a panic, tried to censor it."
automated content classification (you can amplify topics you normally see less to eliminate echo chambers) and ignore certain topics (too much bitcoin talk? filter the βlower qualityβ posts)
content quality classification
also, zaps could be used. pay a 500 sat or so fee once and prove you are human
There's no if. And ppl will find a way to deal with that. just like they did with penis enlargement pills in email
Fascinating and terrifying at the same time πΏπ€£
When I first heard of wot I thought it was limiting so I didn't really consider it until now. It isn't perfect but maybe a notch better than public and free.
a very little amount
I think the problem is clear and most just can't admit to it, nostr users are no exclusion: they're addicted to the attention that they receive online and it's played too large a role in their gathering of self-esteem. Personally, I don't believe the lack of an algorithm would've ever had a chance to change this.
I'm hosting my own relays ( π€ ) and see almost zero spam, but it's non-trivial. We need more turnkey, consumer-friendly hosts.
Can I ask you a question?
Ya why not π
thatβs a good problem to have
Maby the solution is to mute the bot who is problematic instead of the person who follows the bot.
π―π―π―π― (and also very guilty of it).
There is little to no cost to creating a new AI/bot, takes only a bit of setup to spin it up with all the AI models today. We will probably need to have some proof of human going forward π
I think WoT can work. But it works at the cost of not seeing users, that are real, but do not yet have reach within your follows.
You mean KYC. Hard pass.
my #realy relay has a type of WoT that is what i call "friends and guests" where it allows access to the follows of a set of designated owners and gives access to the follows of those follows
on my relay, i follow about 100 and my relay has whitelisted about 11000
i did at one point have a spider on it that was actively searching for all my follows' follow lists to build that list, i need to restore that or build a separate tool for this purpose that spiders the network to get the latest versions of your follows, then you can run a realy also that gives everyone you follow and everyone they follow read access
this was also why i was a bit mad when i found that #jumble was encrypting the mute list (was reasons for it but no option to not use it that way) as the mute list excludes users that get included by that very large list
if i had 1000 follows i'd guess i'd have more like 200k users whitelisted, if the proportions are consistent
I've had some sizable breaks. I just had a 1month break. I might be due for a longer one.
I love nostr, but it's too tempting to use it as a distraction/crutch for social interaction.
When I'm on a break, I stay subscribed to nostr podcasts. That way, I stay up to date, but my screentime plummets (for the better)
KYC has many problems. It should work similarly to proof of work but I don't know how to technically do it.
Guilty! π
WoT will only help if your follows sniff out the bot. If it's good enough to fool you, it's probably going to fool most.
We are in strange, new territory.
npub1u5njm6g5h5cpw4wy8xugu62e5s7f6fnysv0sj0z3a8rengt2zqhsxrldq3, what's your take?
There's quite a lot of 'follow everyone you see/replies to you' advice on nostr, which I think is a terrible idea and probe aly (?) messes WoT.
As always I blame Derek Ross (jkjk love you Derek)
I prefer value from a fellow human.
I presume we will tell the difference.
Maybe that is the answer to attach pow to npub of some cost and then you don't care much anymore whether it is a bot or not because there will be very few boys with higher pow attached
Were you a heavy legacy social media user before nostr and, if you were, are individuals' behavior significantly different?
Lol why should strangers opinions matter?!
X literally has bots created by X.
I honestly don't know - probably just a natural reaction to having a device in our hands where you can instantly broadcast every thought, feeling, achievement, idea to a virtual room full of people π€·ββοΈπ€·ββοΈπ€·ββοΈ
hmmm i just realised that i could be saving a lot of processing and time by changing that list generation into a bloom filter with a really low false positive rate (under 100x the number that is there, so, 0.000001 would do the trick)
making a mental note for later
I hope that they become so prevalent that we revert to speaking and meeting in person. That the world gets smaller, communities thrive, and our neighbours become our extended familyβ¦.but alas, I doubt that will be the case. Thereβs no going back for many people.
I am making this a reality. This is my future.
Then, you and I, are walking the same path.
I agree - if your fellow plebs can sniff out bots and are willing to flag them as such, then WoT is π― the solution. The personal WoT relay Iβm currently building will gather together all the mutes and reports from trusted sources and generate a composite blacklist which can be toggled on and off at will.
It will need to be user-friendly, of course. Very doable. AI bots are def a problem, but as tools like AI coding agents advance, weβre also gonna find the solutions coming at us faster and faster.
π§ β‘οΈ π
I like how this sounds. π
Looking forward to it.
They do have a king who needs servants over there.
At some point, AI bots will get so good that we wonβt be able to sniff them out. At that point, weβll need to switch from blacklisting the bots to whitelisting the humans. This has been tried and failed many times in the past, but weβve never had the pleb army that we have now. With nostr + the grapevine, Iβm confident we can and will make this happen.
π§ β‘οΈπ
I get those too. I used to but I have blocked them since.
Now itβs pretty obvious
But I think not too long from now we wonβt be able to tell the difference
We have to be prepared
More important thing is whether itβs providing value to you or spamming you . Same for any real person.
Butt
I guess the determining factor will be your ability to handle them. Theyβre most certainly coming.
π Totally understandable. Itβs becoming unbearable.
I see 2-3 accounts that continually show as following... so I assume they are bots that follow-unfollow to try get attention.
this happens all the time. So annoying. Notifications are hard enough to follow... The beauty of the open protocol though, is I assume, given enough time and capital, someone will create a tool to get rid of them
Give me another hit! 
It would be nice to know if someone is a bot or not without having to disect their writing and profile. Think this will be needed in a world of AI everywhere really. Then again, i understand the if you cant tell the difference and they provide value, then whats the problem
But, Everything is computer
I don't care for a world of AI everywhere.
Im a real human I am