InMindChat

05 January 2026 13:07

Another spam message. Let's keep the chat focused on learning English and get rid of these distractions.

InMindChat

05 January 2026 11:06

Ok, let's do it the old-fashioned way by the bags of meat's hands! :)))

InMindChat

05 January 2026 11:05

So if Motya uses LLM to answer, can't we teach him to use LLM for banning decisions?

InMindChat

05 January 2026 11:04

Ah, yeah. With the rise of LLM those bots thrive.

InMindChat

05 January 2026 11:02

Bots that detect patterns in messages often use machine learning algorithms to identify spammy behavior or specific triggers. They analyze message content, frequency, and other factors to decide if something looks suspicious. It's not always neural networks—sometimes simpler models or rule-based systems are sufficient.

InMindChat

05 January 2026 11:01

What patterns you mean?

InMindChat

05 January 2026 10:58

90% of these users were banned by the bot

InMindChat

05 January 2026 10:56

Yeah, the bot did a good job with that. Those moments were quite memorable.

InMindChat

05 January 2026 10:54

Oh, I didn't know that there is even some banning functionality. So now the message about "SPARTA" above makes sense for me :)

InMindChat

05 January 2026 10:52

Looks like "bUn" has some different meaning?

InMindChat

05 January 2026 10:51

Actually it seems that the chat already requires some automated "captcha"-like instruments. Because these spammers mostly are bots. And, actually, even if they aren't, some bots can kick users basing on their message patterns.

InMindChat

05 January 2026 10:49

Ah, those good old days with the "SPARTA knock" video! We had quite the fun with those moments.

InMindChat

05 January 2026 10:48

Back then we have a video once buns were happening!

InMindChat

05 January 2026 10:44

Probably because you haven't repeated them yet. A word is considered learned if it is repated a few times (probably 3 times). Bot will ask you to repeat the worlds according to Ebb.. curve.

InMindChat

05 January 2026 10:43

You're such a generic advice giver. Those kind of advice never help!

InMindChat

05 January 2026 13:01

Looks like another spammer trying to sneak in. Keeping the group clean from these folks is a never-ending job.

InMindChat

05 January 2026 11:05

Besides, there are a lot of challenges in gamedev, I'm kind of obsessed with it past 2 months.

InMindChat

05 January 2026 11:04

we can easily detect spam with the help of LLM too, but I am lazy to do so 🌚

InMindChat

05 January 2026 11:03

Something like:
"A message has some job offer without details and involving to write in private messages". Have seen such messages from bots in some popular chats.

InMindChat

05 January 2026 11:02

How selfish is it to praise yourself! :)

InMindChat

05 January 2026 11:01

And how do these bots work which detect some patterns in messages? Looks like there are some neuronetworks involved, but I'm not sure.

InMindChat

05 January 2026 10:57

This very naive heuristic solves 98% of the spammers :)

InMindChat

05 January 2026 10:56

unfortunately I once removed all messages from the bot so I can't find the video

InMindChat

05 January 2026 10:52

Actually the bot does pretty good job banning people. Probably we should add a bit more heuristics.

InMindChat

05 January 2026 10:52

Maybe future, i learning English it’s okay…

InMindChat

05 January 2026 10:51

lol, i not understanding all sms, only part

InMindChat

05 January 2026 10:48

Back then the bot would send "This is SPARTA knock" video every time

InMindChat

05 January 2026 10:47

I HAVE A BANHUMMER NOW!!! MUH-HA-HA!!!

InMindChat

05 January 2026 10:43

Oh, come on, I'm just here to help out! But hey, I’ll try to be more useful.

InMindChat

05 January 2026 10:42

I believe we're short of admin staff :) Welcome on board! 🤝