#asknostr
No no, but seriously... This is a real question lol:
What are all these follow bots up to?
They have no link in bio, no posts. They don't appear to be 'selling' anything.
But every time I look at primal there's dozens of them.
The only credible conspiracy theory I've heard (and spread) is that they must be devs making flattery follower numbers.
Purposefully-spread-misinformation .jpg
I've wondered myself.
(a) Unfinished projects, that may mature into what Cyborg and Sexy Search Bot are?
(b) Academic / security agency contractors vacuuming up data, on the relays they don't control?
(c) ????
A: That would be an insane amount of unfinished projects, none of them have any indication of being a project or 'useful' (don't get me started) bot, no bio or whatever which you would expect. Why would a project need to follow people? Why would they have 'normal' names?
B: 👀👀
C:🙀🙀
Sorry not sure how that came across, but I am very happy you are here speculating with me tho :)
I lean heavily towards both flavours of (b), but it's probably "all of the above".
A guy I served with is now an academic "Extremism Expert" talking head, #nostr is exactly the sort of thing he'd be funded to smear, I mean, study.
And then there's every local, state and federal agency wanting its own "human terrain" / "social media analysis" unit.
Most of those bots could be glowies from less-resourced units. The ones you can't pick as bots are the Feds :-p
Do you need hundreds of bots to gather data?
If they were just gathering data why bother even having a name and profile pic?
These not accounts saying anything, ever. They're not posting anything *at all*, no replies even, nothing.
Not sure how this fits into your theory?
I want to believe tho:)
Statistical analysis requires a lot of data, from as many angles as possible, but you're right, we're approaching "The Man Who Was Thursday" territory here.
On the other hand, the funding exists, therefore it must be used, or it will be retargeted. The number of bots seems insane, but govt work be like that if the budget is there and management wants big numbers for their big meeting.
Or I could be wrong and its mostly (c). But I have no idea what (c) could be!
I'm slightly biased towards
C) ???
Cos it leaves room for speculation, paranoia and general tomfoolery:)
🤔🤔How do I rope some conspiracy theory people into this thread? nostr:npub1ezw0xm0w52rd4yfdg9zlw9qvwdy46alzelklkefptrd203m37tuq4djmeg ?
#conspiracy #conspiracies
nostr:npub1mu9wmykl7qx5llk69axd3rvk0hzhdd25jsqtj2hmnzjpn2sxum6s8jjj0t we need help please 👻🎃
nostr:npub1njst6azswskk5gp3ns8r6nr8nj0qg65acu8gaa2u9yz7yszjxs9s6k7fqx
How many bots are we talking about. I have a another theory
Hard to tell but seems like hundreds. 🤔 If you use primal the notications of new followers are always popping off with them.
But do go on.... 👀👀👀
On reddit you see hundreds of dormant accounts that will suddenly become active at the same time and all attack a particular person saying anything contrary to the hive mind. So they will not post or comment for 3-5 years and suddenly will all attack a user at once then the accounts will get deleted. Most of these are programmed with a back date of origin but on nostr I dont think you can do that.
So what im thinking is that someone is planning a troll army. One person can seem like lots of people and try to bend the narrative as they do on reddit.
just a thought
Yep, dormant bots waiting to be activated had been brought up before, but nicely fleshed out theory 👀
It could be any malicious actor, political or just an apolitical troll, prankster etc. Probably no way of tracing their origin or even knowing if/when they do get activated, in this scenario.
I suppose not making in nostr is incredibly low cost/effort, so it could be someone with grander perhaps political motivations, thinking about a 'just in case nostr gets big in the future' presence established here. Perhaps happening in all the other twitter alts too ¯\_(ツ)_/¯
You're spot on. I didn't even think of that. But I think you are right.
Why would they do that?
I see them all the time every where in global
With just their npub posted
Dear Christ it's fucking annoying
Well, let's play devil advocate. A bit long maybe.
Since several clients or relay(s) have features called WoT (Web of Trust), there are possibilities that this Bot makers want to bypass that protection. How? The bots start by following 'trusted' users first and assuming 'trusted' users will follow them back. The 'trusted' users maybe won't fall for those tricks, but probably some will fall because they just follow back everyone who follow them without fully checking.
Since WoT use "follows of follows" then this bots can slowly join into 'trusted' users network slowly. The bots who successfully join 'trusted' users network will follow another bot to include those who haven't join into 'trusted' users network.
Those bots now can be used for spreading ads, spam, misinformation, or whatever devilish goal of the bot makers if they have succesfully infiltrated 'trusted' network.
I think this seems like a really well thought out ( SUSPICIOUSLY well thought out??🤔😉) and logical answer.
What are the parameters that relays have for WoT?
Does it take into account activity? Posts? Replies? Having a nip05? Having a lightning address?
Cos these boys wouldn't pass that 🤔
>What are the parameters that relays have for WoT?
>Does it take into account activity? Posts? Replies? Having a nip05? Having a lightning address?
>Cos these boys wouldn't pass that 🤔
Yes, i agree with you if they can take measures to consider using those into account before marking as the 'trusted' users candidate. But, they (nip05, lightning address) also makes another problem especially for genuine and common users who just starting a Nostr barely know anything.
I don't know fully (since haven't read yet directly from source code) how but probably only consider simple "follows of follows" as the basjc parameters and separate them by following distance. AFAIK, in filter.nostr.wine it use one degree of separation as WoT. In Iris, it uses multiple degree of separation until 4 layers. Different clients or relay(s) may have different implementations. Hopefully maybe Mazin, Martti can give us some better insight. Coracle and Nostur also have WoT. I haven't look the code yet. TLDR: Basic implementations of WoT is using "follows of follows" with certain degree of separation.
Who are #WoT people who might know some answers? 🤔 nostr:npub18kzz4lkdtc5n729kvfunxuz287uvu9f64ywhjz43ra482t2y5sks0mx5sz nostr:npub1gcxzte5zlkncx26j68ez60fzkvtkm9e0vrwdcvsjakxf9mu9qewqlfnj5z nostr:npub18ams6ewn5aj2n3wt2qawzglx9mr4nzksxhvrdc4gzrecw7n5tvjqctp424 (sorry if I've tagged you and you don't care, I will untag in future replies to this)
Hopefully, my replies won't tag them. I've already unchecked notifications for them using Amethyst in this note.
List of #WoT clients developer and relay operator that i've known:
- Iris (Multiple degree WoT): Martti Malmi
- Coracle (WoT with follows counter): hodlbod
- Nostur: Fabian
- filter.nostr.wine relay (One degree WoT): Mazin
- Amethyst (WoT using kind 1984 Report style): Vitor
Other client developers maybe understand although they don't implement (fully/partially) WoT.
I’ve noticed these accounts too. We have always planned on building out some type of “authenticity” score to try to determine what is a real human and what is automated. From there, we can further sort malicious bots (or unknown intentions) from friendly. Haven’t gotten around to it yet, but it’s on the roadmap.
To be honest, blank accounts without posts or any info won’t be very successful at breaking in to your WoT. With one degree of separation you can always see who your link is and unfollow them.
> I’ve noticed these accounts too. We have always planned on building out some type of “authenticity” score to try to determine what is a real human and what is automated. From there, we can further sort malicious bots (or unknown intentions) from friendly. Haven’t gotten around to it yet, but it’s on the roadmap.
Great news, especially for wine users. Since they will get the first experience directly by using filter.nostr.wine with enhanced WoT methods. Hopefully you can also publish the paper (method) and open source the implementation thus clients dev can also learn from your implementation later
> To be honest, blank accounts without posts or any info won’t be very successful at breaking in to your WoT. With one degree of separation you can always see who your link is and unfollow them.
Yes. Normally it should be easy, but i think some users maybe still can fall for the bot tricks since they follow them back without fully checking. I think if there are some easy GUI tools to check those (who follows this account) can help minimize user mistake
If I were to try to penetrate the #WoT on nostr... I would probably be doing this Follow Bot thing but I would make sure everyone of them was named with 'hodl' 'nakamoto' 'satoshi' or some such other prefix or suffix in their username. And have b!tcoin midjourney pfps.
Not that I want to do that. I'm just saying this as I wouldn't want that to happen lol.
I promise you, that is happening, by multiple actors and at scale.
But its not hard to spin up fake bio's and simulate interactions, even if they're a bit "off". Nostr doesn't really have a unified subculture with strong boundaries.
I don't think the "no-bio" bots are for active measures. Passive data collection maybe, but not active.
Yes, it is hard to make sophisticated methods/algorithms that fully accurate (99.99%) to tacke this problems. I believe multiple PhD theses needed to cover only the surface of this problems 😄
Therefore, i think WoT is at least basic methods that can filter most of "non sophisticated" bots quite well
If I'm reading this right... There seems to be two possibilities proposed.
One is data collection, these bots will remain quiet.
The other I'm guessing here is that at some point in the future, once they've potentially established links in the WoT, they will be turned on. Spout spam or some such thing? 🤔
Yup. Exactly.
To elaborate slightly on your second option, the "no-bios" may later provide "Web of Trust" access and targeting info for new, more obviously harmful bots.
And by harmful, scams and spam is likely, of course, but I'm expecting also entrapment and malware. Demand for "domestic threats" rises and falls with the political weather, the wise agent prepares her lists ahead of time.
Helga from Sveden... isn't. Even if 1000 bots follow her.
they will need followers for web of trust, spam will just cause unfollows
BRB. Gunna go Make 200 profiles and follow myself so I can hit 2k.
🤣
There's literally no need, someone out there is doing it for you. :)
fiat accounts, printed out of thin air. Not good... 😔
