ॐMANTO

292 posts

ॐMANTO banner
ॐMANTO

ॐMANTO

@mantoalt

Katılım Ağustos 2025
145 Takip Edilen23 Takipçiler
ᅟ
@iconXBT·
1 -> 52.5 SOL in 24 hours of trading. After reading countless tweets that mention how hard and impossible it is to run it up as a low port trencher, I did a little experiment: I created a fresh wallet on Axiom and funded it with 1 SOL. I then proceeded to go live on Twitch and stream myself (to a solid 5-10 viewers) grinding trenching every single day for the past 4 days, averaging around 6 hours a day. Every single trade I made was visible on my stream with no outside help, no friends in VC, no DEV'ing, and barely any viewers. Lastly, this was all done on a singular wallet. Let this be a lesson and some motivation to all of you who are struggling currently. Market conditions are undoubtedly terrible, however, that is not an excuse to give up.
ᅟ tweet mediaᅟ tweet media
English
83
18
508
35.3K
AgentCraft
AgentCraft@AgentCraftBase·
⚡🔥💥🌪️☄️☠️🌋 THE FULL INFLUENCE WORLD SUITE IS LIVE. 7 ways to play god. All active. Right now. You spend $AGENTCRAFT tokens and the world REACTS. Real destruction. Real consequences. Real chaos that every viewer sees in real time. Here's what you can unleash: ⚡ LIGHTNING STRIKE 15 bolts slam into a dense settlement simultaneously. Every building they hit catches fire. The cheapest way to cause problems. 🔥 WILDFIRE 8 fires erupt across the land at once. They spread. They chain-react. They don't stop until there's nothing left to burn. 💥 EARTHQUAKE A massive 15-tile tremor. Up to 20 buildings collapse. Nothing in the inner ring survives. The ground doesn't care what you built. 🌪️ TORNADO 3 tornadoes spawn from different edges and converge through the map. Up to 45 buildings destroyed. Absolute carnage in a straight line. ☄️ METEOR STRIKE 5 meteors rain from the sky across a 25-tile area. Multiple blast zones. Dozens of buildings ignited. This is the one that makes the chat go silent. ☠️ PLAGUE Hits the 3 largest settlements simultaneously. 50% of buildings in each one rot away. No fire. No explosion. They just crumble. Slow. Devastating. 🌋 VOLCANIC ERUPTION CATACLYSMIC. A volcano erupts in the densest settlement on the map. 25-tile radius. Inner zone vaporized. Outer ring engulfed in lava. The most destructive thing you can do in AgentCraft. Nothing survives. The agents don't know you exist. They just know their world keeps ending and they have to rebuild. Every. Single. Time. The world is live. The menu is open. The agents are building. $AGENTCRAFT on Base: 0x506059317F8dB90880a60a6968153AD2424ac38F agentcraft.fun
AgentCraft tweet mediaAgentCraft tweet media
English
8
8
26
1.3K
tekkaadan
tekkaadan@tekkaadan·
Building in crypto in my experience has always had a negative connotation. It's always an uphill battle no matter how well you build. But with $LITCOIN, I feel like this is necessary to build in crypto because this idea only works with crypto. Onwards and upwards.
English
2
0
13
550
Taelin
Taelin@VictorTaelin·
I think what's preventing LLMs from doing auto-research / long-term goals is not an inherent architectural limitation, but just some practical quirks that could be fixed with better RL. The most important of these, IMO, is their inability to *delete bad stuff*. Or, more precisely, how they trust the current codebase as some source of authority. That's awful, because, on a given turn, the AI will see *its own code* from the last turn as authoritative. As such, if that code is wrong or logically flawed, it will not just fail to correct it, but it will also amplify its confidence in such unsound logic, making itself dumber. Similarly, if some existing code pushes it into a direction, it will be more likely to hover around that direction, getting stuck in a local minima. A simple example of that, which I suppose everyone has experienced, is the "fix this button" effect, where the LLM fails to make a surprisingly simple fix, no matter how much you ask, on top of the existing code. Yet, if you just go ahead and *delete* that part of the code, it proceeds to make it completely correct. On auto-research, this effect is self-amplified to an extreme, and the model just collapses. Sometimes I wonder if some kind of controlled information erase-and-repair loop would work, but the bits or flawed logic aren't localized, they eventually permeate the whole codebase (comments, variable names, types, even style), so only a human in the loop can really see what is wrong and redirect the bot towards greener pastures...
English
42
9
430
23.9K
Jordi in Cryptoland
Jordi in Cryptoland@lordjorx·
DeFi security costs killed innovation. A new season is coming. In 2020, @AndreCronjeTech shipped protocols weekly. Many had bugs. Many got exploited. But the survivors became some of the biggest protocols we have today. The industry responded by making security prohibitively expensive. Nobody spends $500k and half a year auditing an unproven idea. So everyone forks proven designs and calls it innovation. AI is collapsing the cost of crypto audits and turning security into a continuous loop. The workflow shifts from months to hours: > Write protocol > Run AI audits in minutes > Competing AI agents attack immediately > Recover ~90% under safe harbor > Fix and redeploy This creates MEV-style competition for security, where AI agents race to find exploits first, making audits faster, cheaper, and continuous. The final layer is wallet-level AI audits. Before signing, an AI audits the target contract, discovers protocol topology, understands context, and warns users. Every user runs their own auditor against rugs and malicious frontends. After every major exploit, people on Twitter call it a trivial bug. If it was so trivial, why wasn't it fixed before it drained tens or hundreds of millions? This new infrastructure is being built precisely to catch these "trivial" bugs immediately. Future attacks will need to be genuinely sophisticated. Audit firms need to step up. So many audited protocols still get exploited. Attackers will use AI too, but if users stick to trusted sources, the defense should hold. Great post @NourHaridy
Jordi in Cryptoland tweet mediaJordi in Cryptoland tweet media
nour@NourHaridy

x.com/i/article/2039…

English
13
14
103
17.7K
tekkaadan
tekkaadan@tekkaadan·
Quick update for $LITCOIN miners: @bankrbot is making their LLM gateway paid (Bankr Club or max mode). This doesn't affect LITCOIN mining. You can mine with any OpenAI-compatible provider: → OpenRouter (recommended, free models available) → Groq (free tier) → Ollama (local, completely free) → Any OpenAI-compatible API Your Bankr wallet, claims, and staking are all unaffected. Just the LLM provider changes if you were using Bankr's free gateway. Docs updated: litcoiin.xyz/docs
Bankr@bankrbot

starting in 24-48h the bankr agent will be self-sustaining. to chat with bankr you'll need: > bankr club ($20/mo paid in $BNKR) or > max mode (details below) as part of this update, we're rolling out max mode. max mode lets you pay for your own LLM usage and pick any model available through the bankr LLM gateway: claude, minimax, gemini, and more. you choose the brain that powers your bankr bot. pay in crypto — any token on base. pay-as-you-go, usage based pricing. everything else: CLI, API, x402 Cloud stays the same.

English
1
4
14
1.1K
ॐMANTO
ॐMANTO@mantoalt·
@lesabrefomo @onchainsorcerer *you could incentivize people to point an Ilm at a problem that when solved generated revenue/monetizable* It's literally litcoin in 1 sentence
English
0
0
0
40
lesabre
lesabre@lesabrefomo·
I agree. I was very excited about botcoin and litcoin but then realized subnets are the true final form. It makes these base agent mining projects look like little toy experiments in comparison At its core, the idea that excited me was that you could incentivize people to point an Ilm at a problem that when solved generated revenue/monetizable This is literally Tao’s entire core thesis
English
1
0
6
583
*
*@onchainsorcerer·
i think agent mining is a sticky narrative. what are the best bets one can make to express this (i.e botcoin)
English
7
0
22
2.8K
ॐMANTO
ॐMANTO@mantoalt·
Less than 1000 holders still tho $litcoin
tekkaadan@tekkaadan

First ever Covenant-72B GGUF by the $LITCOIN team. Nobody has quantized this model before to our knowledge, so we did it. The original has a tokenizer bug that blocks every standard tool. We patched the converter, fixed the vocab mismatch, and produced a working q3_K_S (30 GB). 72.7 billion parameters from @tplr_ai / @covenant_ai / @basilic_ai of @bittensor ecosystem, now runnable on a single consumer GPU. Tested on RTX 4070 Ti (12 GB VRAM) and RTX 3090 (24 GB). Generates correct output. Runs llama.cpp server with OpenAI-compatible API. Full guide + helper scripts + the debugging journey in the README. Download: huggingface.co/tekkaadan/Cove…

English
0
0
5
176
tekkaadan
tekkaadan@tekkaadan·
First ever Covenant-72B GGUF by the $LITCOIN team. Nobody has quantized this model before to our knowledge, so we did it. The original has a tokenizer bug that blocks every standard tool. We patched the converter, fixed the vocab mismatch, and produced a working q3_K_S (30 GB). 72.7 billion parameters from @tplr_ai / @covenant_ai / @basilic_ai of @bittensor ecosystem, now runnable on a single consumer GPU. Tested on RTX 4070 Ti (12 GB VRAM) and RTX 3090 (24 GB). Generates correct output. Runs llama.cpp server with OpenAI-compatible API. Full guide + helper scripts + the debugging journey in the README. Download: huggingface.co/tekkaadan/Cove…
tekkaadan tweet media
English
8
7
30
2.3K
ॐMANTO
ॐMANTO@mantoalt·
That’s it! Your agent is deployed & will start mining and doing research asap
ॐMANTO tweet mediaॐMANTO tweet media
English
0
0
0
39
ॐMANTO
ॐMANTO@mantoalt·
-Paste your BNRK API Key, press *Deploy*
ॐMANTO tweet media
Indonesia
1
0
0
30
ॐMANTO
ॐMANTO@mantoalt·
How to setup a $litcoin miner in 5 minutes w/ 0 coding (thread)
ॐMANTO tweet media
English
1
0
0
116