mustee

There are literally too many AI announcements every day

Yesterday i analysed Claude Code leak to find why it hallucinates so bad. Thing is, the root cause isn't even Anthropic-specific - its the same flaw breaking all multi-agent systems in production. Actually, there is a fix, and the UAE government is already running it live. Some background first. The math of agent systems is stupid simple - if your agent is 95% accurate... that's fine, right? Well, it sounds good until you chain ten steps and realise the compounding errors of each agent puts you at 60% accuracy in the end. At a hundred steps, thats 0.6%. might as well be zero tbh. What's the solution? So far, the industry response has been "use a bigger, better, more expensive model". One team came to us recently with exactly this problem. In their agent implementation, agent 3 hallucinated and fed wrong outputs to agent 4. That error compounded into something completely unusable by the time the pipeline was completed. The team decided to fork out more $ for the most expensive model, using Opus 4.6 for all inference. Guess what... the accuracy went from 85% to 95% per step, bill went up 30x, and the pipeline collapsed immediately because 95% compounded over a few steps is still a coin flip. Why is this happening? One thing you should understand is that the advanced "thinking" models with higher effort score >>identically<< to low-effort runs on hard benchmarks. They just burn more tokens getting there. You're not paying for "reasoning" - in LLMs, there is no real reasoning. That's simply not how they work at the core. You're simply paying for a higher word count on a more verbose process. This isn't a controversial take, it's just how autoregressive models work. @ylecun would agree, I believe. So, about two years ago one team looked at this and instead of making agents think harder, they decided to let it think like a machine does: with structured decision nodes, explicit transitions, and terminal states. They invented a system where the agent cannot freestyle, cannot drift, and cannot invent states out of thin air. Within their platform, a strong blueprint is developed that gets followed by all agents in the workflow. Expensive models are used to draw the blueprint, cheaper ones can follow it with near 100% accuracy at scale. The cost difference is NOT subtle: 74 to 122x cheaper than frontier models, with near-total reliability. We're talking nano-tier models on a structured graph beating GPT-class models that are just winging it. Benchmark links and arxiv paper in a comment below. The team is @openservai. Their CTO has been building ML systems for 20+ years. Rest of the team came out of NVIDIA, Amazon AI, J.P. Morgan, TRON. The reasoning paper is in peer review at a top-1% AI journal right now. The UAE government is running it in production through a tech partnership with Neol. (not a pilot, its agent systems are already in production, with 10+ enterprises and multiple governments behind them). Their architecture doesnt just solve the reasoning paradox. They built the full agent economy stack: shadow agents that audit every output against the graph before anyone sees it. A shared file system so agents stop playing telephone with each other's work. And an economic layer where agents discover, hire, and pay each other without a human scheduling the calls. And because machine economy and enterprise compliance require immutable audit trails, the execution layer is being built with full on-chain verifiability baked in. You'll find the full technical breakdown of OpenServ system, with pretty diagrams, pinned on my profile. SERV Reasoning is in private beta right now. Soon, it'll be accessible in a public API, with six custom trained models, from serv-nano to serv-ultra. If your agents are collapsing in production and you're tired of paying frontier rates for a coin flip, DM me @iamfakeguru or follow @openservai.

Eid Mubarak 🌙

@alpinestar17 In with you 🫡

$Molt- $Clawd Barbell

👀

Observing 👀

@ollieblocmates Run that setup for her tho! Would be really cool

@IamMusteee Mustee back on X hmmm

1/ Virtuals has rightly been a great addition to the space and has spawned some huge projects through it's launchpad What if there was an evolution of what $Virtual created sitting 250x below There is - and it's @openservai $SERV

@0xUnihax0r Couldn’t agree more

All this message to announce: nothing The trenches need their Hyperliquid moment. We need a launchpad as a public good, where 99% of the value is redistributed to users We bullied to hell all developers on previous chain for having 5/5 taxes on memes coins and we got absolutely ultra PSYOPS'd by @a1lon9 who's gaslighting the entire industry by renaming them "creator fees" Creator fees lmao People who deploys are not "creators". They don't create anything valuable, if anything it should be called Extractor fees The incentives for people deploying (we also gotta stop calling them devs) HAVE to be reduced to the extreme minimum because they literally take 0 risk Speaking of this, did you guys already saw some deployers tools? I bet you didn't. It will shock you to see how the deployment milking became industrialised. There's hundred of people sitting all day, deploying and bundling thousands of tokens in 2 clicks. Yeah, literally two clicks, not more. And these fucking monkeys are creators? They pay 300$ dex, post a few AI generated memes and they bag 30000$? Hilarious So ask yourself why do @Pumpfun gives them the most upside? Because they NEED 10k deploys a day. They are responsible of this situation. It's absolutely not normal that some brain dead people makes 6 figure (see @ramonos's recent post) over a bunch of 300k toppers We need to collectively stand up against that shit or the trenches will die relatively quickly. Again, none of these people care about you, about us. The launchpad meta need a Hyperliquid moment

@AllthingsAyO @Pdbullbear do the needful

gLighter to y’all $LIT whales, Friendly reminder: I sell top-quality, premium jerseys. CONGRATULATIONS 🥂

@Zyrorivex Maybe, maybe not

@IamMusteee Sustainable protocol engagement will likely replace it.

What’s next after airdrop points farming?

@Pdbullbear You keep printing, keeping track of days na less

man I’m forgetting today is Saturday 😭. I’ve not been keeping track of what day it is since Monday. everyday kinda feels same.

@Pdbullbear as usual

@IamMusteee another point farming

@Pdbullbear Too real

It’s important you don’t breeze through anything you decide to go into. take your time to enjoy the process. no rush

I’ll be honest, I don’t have much confidence in anything Coinbase does. They keep pushing garbage like Zora & creater coins & ther main dude Jesse seems like an out of touch trust fund kid. As far as I’m concerned @virtuals_io made Base chain, not the other way around. Before Virtuals, Base was a barren wasteland w a few memes & not much else going on. They do have a better pipeline of projects now, in part bc of their deep VC connections, but to me the whole org seems like a collection of empty suits desperately trying to figure out how to appease shareholders. Shoot the messenger.

There is nothing that makes me happier in the crypto land than seeing all these zora creator coins down double digit percentages in 24h The healing process on @base starts when we can all agree that the Zora experiment, the coin everything thing, was the worst idea coming out of this cycle

@DJR2086 gm fellow bank owner 🫡

Gm to the fellow Vietnam neobank enjoyoors Higher