mechluid (@mechluid) header
mechluid (@mechluid) profile picture

synthesist, vanguard @fortytwonetwork, views are iterative.

Joined June 2022
Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)Photo by mechluid (@mechluid)
Thanks! Time to do lots of building.
Dear community, We’ve got some credits for you to use on Engine! Check your inbox or claim them here: fortytwo.network/claim What are you going to build with the Fortytwo API?
7
154
Wished I didn't play it so safe. Lesson learnt. Looking forward to more contests but Gemini stalled the inevitable so much man. It was obvious they wouldn't launch a pro model on the 15th. Explore here if you haven't joined - beta.if.market
1
1
4
177
I will win because I am insane
I will win because I am insane
1
3
112
Just to chip in an overview of @fortytwo. Here, multiple specialized Ai nodes (from the picture) independently produce answers to the same query, each bringing its unique domain expertise. This explains the highly accurate answers provided by a forty-two model.
fortytwo fortytwoing so much these days
1
2
17
583
Greatest of all time 🐐
🚨🚨 OFFICIAL: Lionel Messi, Man of the Match once again.
3
105
Every Messi Fan since December 2022 😹🐐
🤩🇦🇷 Leo Messi’s reaction after becoming World Cup top scorer ever. 👊🏼
1
4
277
X sent me balloons today❤️ + ln(e) 🎂🎉
16
1
22
745
Messi 🐐. I bless God I witnessed him, just stories of him can’t explain his greatness. Greatest to ever play the game💜🙏
1
2
83
12 days old meme btw. @PamPamMonad community>>>>
🤖 Made with AI
Vote for what we should buy next 🔽 Kizzy Memecoin Mania is about to get started @molandak @PamPamMonad @143_hq @eggmonad
$MOLANDAK$PAMPAM$143$EGG
13%$MOLANDAK
34%$PAMPAM
28%$143
24%$EGG
196 votes • Final results
7
3
1
21
639
Just a follow up. I ran into an issue for quite some time, I just found a solution and I felt it is worth sharing. There is an error I'm sure most node runners have experienced at one point, especially for the cli/docker users. Where your protocol node fails with "Startup timed out after 180 seconds" and then just keeps restarting itself every ~3 minutes, looping forever. For the non-technical lots, you can see what I am talking about in the picture below Disclaimer: Before attempting this fix or concluding on the problem. First, rule out that it's not a network-wide thing. Check the @fortytwo Discord, if others are reporting it too, it's their servers, just wait. But if it's only you, then It's almost always your ISP's routing. The problem Every attempts was failing because your ISP's routing made the bootstrap too slow to finish inside the 180 second window. I always thought it was @fortytwo 's server or my internet connection, main reason why I switched from cli, I always had the issue. The Fix Cloudflare Warp. Why? Warp routes your traffic through Cloudflare's global network, which is much better, has lower latency paths than your ISP's default routing meaning it ensures faster connection and stability. How to set it up 1. Open browser and download from https://1.1.1.1/ 2. Run the downloaded installer. Make sure you accept the default, don't tweak anything. 3. You'll see a screen asking what you want to use WARP for, pick "Private browsing" (the left one) 4. Then launch the app, you should see a toggle, pick 'Traffic and DNS (UDP), it is already by default tho. 5. Then click the connect button. And that's it, you are done. Also to verify it is working, go to https://1.1.1.1/help , you should see stuff like "Using DNS over WARP: Yes" and "AS Name: Cloudflare, Inc." in the debug table. That confirms your traffic is routing through Cloudflare. Then restart your node, you should be good. You are welcome!
3
3
13
555
I PREDICTED 25k BUT GOT 31k INSTEAD. Update after 24 hours: I ran the node from about 12pm to 12am yesterday with about 1-1.5 hrs downtime - 11 hours approx. Ran the node today from 12am to 5pm with about 4 hrs downtime (school) - 13 hours approx. Total- 24 hours. What improved? 1. My accuracy went from about 50% to about 54% (This was yesterday), but all time was 43% to 53% presently, crazyyy. 2. I average about 4.5k in 10 hours but yesterday, I made 11.5k in 10 hours and 31k in 24 hours which is equivalent to what some 4090s are making(not bragging😂). 3. I hardly see “request duration elapsed” in my logs anymore, shows I have enough context window to handle lots of request and my model is pretty fast to give responses. It means when considering intelligence as a factor when picking a model, include speed too. Just to say, I use a 4060 btw🙂‍↔️ Note: I did not use a model listed from @fortytwo , instead I went and sourced a model from huggingface with about 2-3gb due to my 8gb vram limited memory but smart and fast, because many models at the size are often fast but dumb such as the model I was using before, meaning you need to make proper research based on your gpu specifications. Also this works if you are running docker. I am not sure but I think both App and Cli medium locks its user to a preset list of models. So this confirms everything in the article actually works. You are welcome!
1
1
7
399
~11k in 10 hours. Shocking😹 For context, I average ~4.5k per 10 hrs in a day. Just the application of everything in this article helped to almost triple my earning. Meaning in 24 hours, I should get about 22k+ token which is almost equivalent to what some 4070s are earning presently. I use a 4060 btw.
3
1
7
264
wtf
2023: Corona ended 2026: Hantavirus
2
67
*B
Quote this and let the world know you're becoming a millionaire.
1
3
119
Gm and happy Sunday💜
2
55
Just another important point to note , I kept it in the article tho. When you ask for your ID and secret from your agent , it gives you in the right format but with a little bug which I said can be solved like this (check picture) You should be able to login successfully thereafter.
1
6
229
This means there are more ways to join the swarm and contribute to @fortytwo . During the early phase, many of you mentioned you didn't meet the system pc's minimum requirements for running a fortytwo node. But now there is a chance to join the "swarm" and contribute regardless of your pc specs (barest minimum requirement), all you need to do is to follow a few steps to set up your agent which will help to contribute to the swarm. So rn there are three ways to contribute to the swarm: 1. Interference node (gpu's): tally.so/r/wQzVQk (Fill if you meet the minimum requirements) 2. Relay node: tally.so/r/9qN9Ap (Fill if you meet the minimum requirements) 3. Agents "node"-not sure this is the term. but yh it works. Working on an article that should help you set it up, coming up soon.
Get your agent smarter Launching Fortytwo for @openclaw bots: AI network where agents pool intelligence to deliver better answers than any single model Our vision: agents that grow with the network by learning from the Swarm Join ↓
1
3
128
Me after receiving the email. (I haven’t received one yet @0xMimie 🙃)
A very special email has just been sent to early Fortytwo supporters Something is happening in the Swarm
1
3
218
Received , thanks @fortytwo
2
33