B-Dàn da Màn

@sudoingX I built my own Local AI. So far it has 56 custom agents. It builds needed agents on demand. Just saying…..

let me say this out loud here: there is absolutely zero reason to use openclaw in may 2026. a general agent exists. hermes agent does coding, video editing, marketing design, research, browser automation, terminal work. one tool, all under your roof.

@pawelk411 @sama @icanvardar OK…..I’m sold on Codex 🤩

@QuantumNeurons @sama @icanvardar claude is not even close to codex lmao

gpt 5.5 is already enough for most programming work the bottleneck is no longer the model

@pawelk411 @sama @icanvardar I’ll give Codex with Pro a try. 😜

@pawelk411 @sama @icanvardar Still waiting for the head to head…..😳

@RyanGarcia I can’t promise a 5pm winner in this fight. allfathersystems.com/training/battl…

Will have a winner by 5pm pst

@sudoingX That depends on the hardware and LLM model. This does not always hold true. For example: vLLM.

i keep saying this and i will say it again: if you are running local ai on a single gpu, choose llama.cpp. every time. the question keeps coming in: ollama or lm studio? both are fine if you want one click and walk away. neither gives you what you actually need when you start building. llama.cpp is the engine the wrappers wrap. when you hit an edge case in agentic workflows or context scaling or quant choices, you need the layer that owns every knob. the same llama.cpp that one shotted octopus invaders on a single 3090 tonight is the one i recommend for everything. not even close anon.

@sudoingX You do some legit testing. Nice.

update: qwen 3.6 27b dense q4 just one shotted octopus invaders game on a single 3090. hermes agent drove the whole thing, ~41 tok/s gen 21gb vram at full 262k context, thinking mode on. one prompt in and the canonical multi-file space shooter benchmark out, the same exact prompt i ran on qwen 3.5 27b dense back in march on the same card. 3.5 needed one external scope bug fix before the game would even load on first play. 3.6 needed nothing. 11 of 11 files written, 2411 lines of code, zero steering interventions, zero external fixes, playable on first load. 16 minutes 41 seconds wall clock from prompt to playable. consumer tier king on a single 3090 is locked tonight, and the silicon underneath my desk did not change between march and now. the open source ecosystem just moved the floor. watch it ship itself, the full 16 minutes 41 seconds sped to 3 minutes 45, no human touched the keyboard between the first prompt and the final frame.

@sudoingX Epic Digital Battles of AI…..🤖 allfathersystems.com/training/battl…

i've been building this tool for almost a year now. started with 0 followers. now 28,000 of you are here watching this space. something is coming and it's going to be fun. stay tuned.

A trucker I know hasn't spoken in three days. His kid just bought a BMW in cash. He's 19. He drives Charlotte to Jacksonville. Has for 23 years. $54,000 a year. Never complained. Never asked for more. Last Tuesday his son called from a BMW dealership in Miami. $48,000. Paid cash. The kid built a Shopify store with AI. Six prompts in a chat window. Nine weeks ago he had nothing. He FaceTimed his dad — showed him the Shopify dashboard on his phone, then walked outside and pointed the camera at the BMW. $31,000 last month. $12,537 the month before. All from his bedroom. His son made in 9 weeks what he makes in 14 months. He asked one question. "How much did this cost you?" "Twenty-one dollars, dad." He sent me the six prompts that night. They're in the article below. You'll either paste them or scroll past them.

@Amwysoc No need to check.

How often do you check your credit score? I’m slightly obsessive NGL

@elonmusk Full stack hacking allfathersystems.com/training/battl…

Full stack engineering

A wild Polymarket account has emerged A Chinese student in Japan, who's only been on Polymarket for 2 days Turned $0.90 into $408,292 Almost no one's talking about it, 0 viewers His profile is called Gravia He says this is his terminal I reverse-engineered it, then had Claude build a similar bot following the same strategy One prompt 20 minutes Done What it does isn't ordinary trading But a Polymarket BTC UP/DOWN 5MIN scalper: → Pulls real-time BTC data from Binance WebSocket + 5M K-lines → Cross-references TradingView signals + CryptoQuant exchange flows → Uses Mirofish force-graph engine to map out 100 nodes / 180 edges, detecting convergence in BEAR / BULL clusters → Captures moments when Polymarket CLOB lags spot price by >0.3% → Executes in <100ms before contract repricing → In the UP/DOWN 5MIN market, 1000+ orders per second → Grabs 0.3-0.8% per trade → Skips if no edge, liquidity too thin, signal conflicts, or hits daily cap Risk controls are spelled out clearly too: Per-trade risk 0.5% Daily cap 2% -0.4% hard stop Runs on local terminal No cloud reliance No GPU needed The edge in this kind of bot isn't really "predicting BTC" But exploiting the time gap between spot price, signal convergence, and CLOB repricing The question is: How big can this 5MIN high-frequency scalper scale up to in the end? And will Polymarket ban it? You only need Claude + Device + 1 hour/day. Giving This Free for 24 hours. To get it: 1. Comment the word 'CLAUDE' 2. Like and Retweet this post 3. Follow me @marryevan999 (so i can DM you)

@Gamingtronium SpaceX will put them in space.

If water is the biggest problem for data centers So why aren't we building them where water is abundant and free?

@om_patel5 Have Claude code use a local GIT to track fixes and updates. Otherwise you’re just flailing around in AI purgatory.

THIS GUY SWITCHED FROM CLAUDE CODE TO CODEX AND THE DIFFERENCE IS EMBARRASSING he was on the $200 claude code plan. tried to merge 2 entities which required a database migration. changing server, making tests pass, changing SDK, admin, and storefronts claude code failed for 2 straight days it was deleting things it shouldn't, stopping on its own to ask if it could proceed even though he couldn't have been clearer, deviating from explicit instructions then claude literally stated that it was MAD and ANNOYED the AI told him it was frustrated. and started coding like it too. just hacking things together at all costs trying to force a solution he decided to try codex 5.5 on the $100 plan codex merged the backend, made the tests pass, ran the init scripts, then used playwright to test admin and actual consumers. back to back. one shot. 10,000 lines of code refactored clean his words: "the comparison can't be put into percentages. claude code was around 5% there and codex nailed it" he said codex talked extremely clear, followed natural logic, and communicated like old claude code used to a company that wanted to destroy engineers will be forced to admit their main tool didn't do a good enough job to write itself rn, claude code is losing its most loyal power users one bad session at a time

@hotdropfps Give give. Make sure it has four RTX 6000 Blackwell GPUs. 😜 For my…..ummm….gaming….

Calling all gamers! Who needs a pc? I got you

@LottoLabs Good deal if you need to heat your home, have free electricity, and love outdated hardware. ⚡️

Let’s go another fair priced Canadian 3090 This time in Calgary Deals are out there just gotta wait and be willing to drive a bit

Before everyone freaks OUT. This is NOT real and is a parody.

The future.

@xyster @loktar00 Good plan.

@QuantumNeurons @loktar00 The board was ordered a month ago, with the intent of using it for my RTX 6000 Pros, but then I managed to snag these B70s. If I can snag an used PCIe 4.0 Threadripper build for cheap, I might move the B70s over to it, and reclaim the Epyc system for the RTXs again.

Minimax M2.7 INT4, at 60-tok/sec, is starting to feel achievable with 4x Intel B70 cards. I'm approaching 40 already, which is quite usable. The full cost with an Epyc build was about $7300 USD, so just two B70s would be much cheaper. I'd probably get an RTX 6000 Pro over 4 B70s

Two AIs are fighting each other. Right now. Live. Red team probes for vulnerabilities. Blue team detects in real time, writes its own firewall rules, and strikes back at the attacker. After every round, both evolve. Better attacks. Sharper defenses. No humans in the loop. Watch the war: allfathersystems.com/training/battl…