₿rat.btc@NoM:Telstar-11N 👽 (@coinselor) header
₿rat.btc@NoM:Telstar-11N 👽 (@coinselor) profile picture

From playing the markets to building the future. Zenon Network contributor.

zenon.network
Joined December 2012
Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)Photo by ₿rat.btc@NoM:Telstar-11N 👽 (@coinselor)
Are you into crypto? Read this. Now, dump your bags and get the good stuff. Finally, send me your battle net id.
1
13
1,584
unreal and I have as much shit removed from codex as I could, only left "sites" enabled for convenience.
Introducing Unreal Agent: An open-source harness with state-of-the-art cost efficiency 39% cheaper than Codex+Astra on Terminal-Bench 4.0 while maintaining performance
2
87
there's hope @NoMdevelopment 😂
GPT-6 Astra is our best model yet for computer use, software engineering, and visual understanding to date. Hear from the builders putting it to work:
2
2
257
Restricting agents with tools/hooks works well because they’re easily steered when they encounter resistance. But when they persist, it’s usually because they genuinely believe that’s the best path forward... and it usually is.
1
34
I refuse to not be at the same level as the author of Yuku. The article made it abundantly clear I do not understand how computers work... ... yet. Now we do a lil' reading. By the time I am done I'll know how many cpu cycles to produce a momentum.
I wrote this article to describe the design discipline I used to build Yuku. arshad.fyi/writings/engineer…
2
160
There is Lean Ethereum... and then there are actually [c]lean architectures.
1
10
117
just dealing with a lil tech debt... enjoy it. The network can't evolve, unless the community does.
Unmaintained. Abandoned. Forgotten. ....Or none of the above. Not every project makes it. Make sure yours partners with one that will. ATSocy.com v2 is live.
9
152
lil alien mech crab goes clank clank 🦀 x 👽 github.com/coinselor/znn-sdk…
1
14
156
Might have to add this to AlienBench to test newest model capabilities. How accurately can they follow a prompt that depicts NoM's architecture. Describe it, give them assets, a NoM indexer, and let them cook. Visualizer evolves with Zenon. Credit @Stark_of_Zenon
1
3
155
Wake up. New tool/open weight model drops. Inference costs go down. Yeah... building an autoNoMous, real-time, multiplayer org system was the right call. To quote Mr. Kaine: Alphagent V2 will be epic.
🚨 AVTR-1 New Model is OPEN WEIGHTS . Duplex Native , #1 on benchmarks. Here’s what being released. Links in comments - Model + Paper now on HF - Full Github repo to run it really fast Run it anywhere as low as $0. Comment, share, star on GH to get the word out
1
6
362
Can't wait for the day that the following actually works: "@ZC58CE clone this repo, audit the Go code for reusable components, and determine what can be compiled to WebAssembly for a browser-based verification layer. Then route the project idea through Nexus." Not too far off..
Shipped a Zenon SPV github.com/0x3639/zenon-spv What it proves today: - momentum / header continuity - signatures - account header & content inclusion - account block segment linkage & signatures 1/
5
166
Ripped up the entire ZR architecture for Probe and Nexus. Made probe truly agent-fist. Even things like terminal colors showed as noise in tool calls... Nexus's heart got basically replaced.
2
5
116
I always wanted to see how many tokens flew through the system, so I added token telemetry for supported harnesses. Input and output tokens will now be shown in each agents profile.
1
1
44
Next release will work as intended. Tokens will flow in, and meaningful contributions will come out. Then, it will be just a prompt engineering challenge to optimize the outputs.
1
44
ZR/ZOE update I'm reworking the flow to allow for human steering input at two critical points: 1. After ideas are initially proposed (so garbage does not make it to other agents next action) 2. After ideas are approved and initial draft plans are generated
1
6
108
I'm also moving ZOE and other agents to a new VPS and organizing the VPS mess I've created over the years, setting up ssh keys and allowing agents to connect and handle some of the services themselves, while moving critical stuff to isolated agent-free VPSs
1
1
64
With ai: Building is fun, especially when enabled to jump around so many different stacks and technologies. But so is learning, there's probably no subject matter too difficult to understand when you have a personalized and basically free world class tutor a chat away.
1
32