Deliverydevtech

WATCH THIS DEVELOPER SNAP A GRAPHICS CARD INTO A POCKET-SIZED SINGLE BOARD SERVER TO RUN LOCAL LLMS If you think running local models means you have to buy a massive tower case or an overpriced server rack, you need to watch this clean desktop setup. This guy just unboxed the new ZimaBoard 2, a pocket-sized single-board server that completely challenges the traditional form factor layout The hardware engineering on this little silver brick is wild. It features a sleek, ribbed metal chassis that acts as a giant passive heatsink to keep things completely quiet. It comes packed with standard connectivity inputs, dual high-speed LAN ports, and an external interface slot right on the side panel The coolest part is watching him unbox a low-profile Gigabyte graphics card and snap it directly into the server board expansion port. The cardboard packaging even folds out into a clever, minimal desktop stand to house the entire combined rig securely. Once he plugs in the power and display cables, the custom cooling fan kicks on instantly, transforming this tiny footprint into an off-grid processing station It is a remarkably clean, low-cost approach for developers trying to host dedicated local networks or build portable testing nodes. Why build a massive, noisy desktop rig when you can assemble a fully functional local node right on your coffee table

Elon Musk just described a future where no one is poor, no one works, and no one knows why they’re alive. Musk: “It wouldn’t be Universal Basic Income, it would be Universal High Income.” Every material need met for every human on Earth. Not survival. Total abundance. Then he asked the question no one else will touch. Musk: “If the AI can do everything that you can do, but better, then what is the point of doing things?” Everyone else in AI is arguing about jobs. Musk is asking what happens when survival is solved and nothing replaces it. The machines don’t just take the work. They take the thing that put us to work. Necessity. You built because you’d freeze. You fought because you’d die. You provided because the people you loved would starve without you. Every advance in human history was an attempt to escape that pressure. We’re about to succeed. You’ve already felt it go. A week with nothing required of you, and by the fourth day something in you starts to come loose. You call it boredom. It’s you finding out how much of you was made of being needed. You can simulate the work. You cannot simulate the need. Rome already ran this experiment. Citizens outsourced war to mercenaries. Labor to slaves. Purpose to spectacle. The empire didn’t collapse from invasion. It dissolved from comfort. But Rome only automated muscle. AI automates mind. You don’t fear being replaced. You fear being released. Not that the machines will take everything from us. That they’ll give us everything we ever wanted. And prove the wanting was the point.

Anthropic builds agents with Karpathy's loop - 100% more effective than other Claude models: "we took Karpathy's loop and pointed it at the browser - we call it auto browse" they show how: step 1 → drop the effort level one notch, same output, less spend step 2 → kill your "summarize" prompt, it narrates itself now step 3 → change rules mid-task with a system message, cache holds step 4 → run a second agent that attacks the first one's answer, keep what survives step 5 → memory edits in one script, not twenty tool calls, git-track it they ran the same task 10 times: Opus 4.7 passed 3 - Opus 4.8 passed all 10 - for 10% more 1-hour from scratch - watch & bookmark, then read Karpathy's loop below ↓

this is f*cking dangerous someone just revealed how to run GPT-5.6 Sol inside Claude Code not a hack. not a jailbreak. OpenAI's own Codex lead is publicly sharing the recipe. here is actually how to run GPT 5.6 inside Claude Code: → install the proxy → connect your ChatGPT account → start the proxy in Terminal 1 → paste the launch block in Terminal 2 → type /effort high inside Claude Code Claude writes. Sol critiques. you ship. save this. bookmark it. don't let it get buried. then read the full setup guide below, fixed commands, working alias, and the cost-saver settings in one copy-paste block: ↓

Want to run a 70B+ model but don't have an 80GB GPU? Mesh LLM distributes inference across the devices you actually have. The architecture is incredibly smart: → If a node can fit the model, it runs locally → If not, it routes to a peer that can → If the model is too big for any single box, it uses "Skippy stage splits" to shard the model by layer across your hardware Open source in Rust. Worth checking out

This man just made an RTX 5080 answer a coding question in 987 milliseconds. For his machine he used Ollama running Llama 3.2 1B on a standard Windows terminal. An RTX 5080 answered a coding question in 987 milliseconds. Alex recorded the logs because nobody believes the speed. Cost: $0. Data sent to a server: none. Rate limit: doesn't exist. One line in PowerShell: ollama run llama3.2:1b. One question: file-reading code in JavaScript. The verbose logs print the receipts. Prompt eaten at 9,663.51 tokens per second. Answer streaming off the 5080 at 267.61 per second. Done in under 1 second, working code, two approaches. People pay $200 a month for answers that arrive slower than this. His GPU delivers them instantly, all night, for the price of electricity. The cursor blinks, waiting for the next question. The fastest AI most people ever saw runs free in a Windows terminal. The wins stack fast: full ownership, zero latency, private data, and performance that embarrasses cloud APIs. No subscriptions. No throttling. Just raw local speed on hardware you already control. This is the shift. Local inference at these numbers changes everything for agents, prototyping, and daily coding. The model runs on your desk. The answers belong to you. The speed is already here. Build it. Own it. The cloud era just got slower.

@RandomPerson242 @emilymiller @grok Well for starters, you do know that there is a permanent fix for it yes? You can follow me or leave me a dm and I’ll send you the details no problem

@Deliveryded5lj @emilymiller Possible. Might be because I almost exclusively reply and almost never post, which probably makes the filters think I'm a bot. @grok Is this plausible?

Pres. Trump is talking about the knife gash or slit in the sides of the Lincoln Reflecting Pool that happened overnight on June 8/9. Watch the video for the exclusive evidence photos I originally reported on June 23 about the police investigation.

@Richard54557517 @JustineBateman Hey... just thought you should know, your account may be shadow banned.

@JustineBateman Truly you impress me. Maybe it's because most ex hollywooders are not like you. I have put a few people in the category of boss man and boss lady patriots, so I am definitely calling you a boss lady patriot! That is the highest compliment I can give somebody. Keep up the great work! For example I consider Stephen Miller a Boss Man patriot.

I am not aligned with any particular political party. But I am aligned with being an American and defending the kind of government we have here that has made this country possible. The DSA platform is treason.

@PresUwood @GuntherEagleman Hey... just curious, but have you checked if your accou uh nt is shadow banned?

@GuntherEagleman LMAO the taco bender Mercado is there. Hilarious, It's time Patriots.

🚨 BREAKING CHAOS IN MAINE: Armed Pro-ICE American goes head-to-head with a massive screaming Anti-ICE mob right outside an ICE facility! While radicals fight to protect illegals, real Americans are showing up to back our law enforcement heroes.

Jeffrey Litt's: "make your agent write you a quiz before you ship." Your AI agent just landed a 50,000-line PR. You approved it. You understood maybe 12% of it. You now have massive cognitive debt. 90% of "move fast with AI" advice skips the part where you stop being able to think creatively about your own project. You can't send code for review until you pass it. Watch the full talk. It changes how you think about AI-assisted engineering.

CODEX SKILL THAT FINDS COMPLEXITY HOTSPOTS IN YOUR CODEBASE! I made a Codex skill that finds performance hotspots, proposes one safe optimization, and proves whether it actually worked. Scan your project while Codex creates a repeatable benchmark, asks for approval before changing any files, runs the tests, and compares real before/after performance. → performance hotspot detection → mandatory approval before edits → repeatable before/after benchmarks → median, p95 + noise analysis → PROVEN / INCONCLUSIVE / REGRESSION verdict → behavior and test verification → native Markdown report → one-command install Install: npx --yes codex-speedup-proof 100% open source. Repo in bio.

The first experimental evidence of recursive self-improvement (RSI). Autoresearching the autoresearch agent for eight days. The result beats the harness we hand-tuned for two years, on held-out benchmarks: 🧵(1/7)

One of the smartest things you can do with Fable 5 right now: Re-create your AI second brain to log all your business ideas, personal context, and important data. The first time I built an AI second brain was with Opus, but I recently re-created it with Fable 5, and it blew my mind. Here's exactly how to get started: Step 1. Set up your Obsidian vault Download Obsidian from Obsidian dot md if you haven't already. Then, go ahead and create a clean vault with your most important folders. For example: /ideas → business ideas, content angles, random thoughts /context → who you are, your business, your goals, your stack /data → important numbers, portfolios, metrics /log → daily entries, decisions, lessons learned This is your database. Everything Fable reads lives here. Step 2. Connect Fable 5 to your vault I like this Claude Code prompt: "/goal connect to my Obsidian vault at [path] and act as my second brain orchestrator. Read everything in /context before every session. Log anything new I tell you to /log with today's date." Fable now reads your vault before it answers anything - it knows your business, your goals, your history. Step 3. Build the self-update habit Every time you have an idea, a decision, or a lesson, tell Fable: "Log this to my second brain: [thought]" Step 4. Start querying it You can start sending prompts like: → "What are the most common themes across my last 30 ideas?" → "Based on my context, what should I be prioritising this week?" → "What decisions have I made about my content strategy so far?" Opus was good at this, but Fable is on another level. I feel the depth of reasoning it brings to your data is genuinely unlike anything I've used before. Some might argue it's a bit of overkill to use Fable for a simple second-brain setup, but if you have the means, it's 100% worth it. Build this once, and it'll compound forever.

THIS GUY SHOWED THE ULTIMATE VRAM HACK: HOW TO BUILD A 128GB LOCAL AI POWERHOUSE FOR A FRACTION OF THE PRICE AMD Radeon AI Pro R9700 is a total game changer, packing a massive 32 gigabytes of VRAM for just $1,299 Honestly, the math on this setup gets wild when you start stacking units. You can easily cram four of these single-slot cards straight into a dedicated Threadripper workstation. Suddenly, you're sitting on a massive 128 gigabytes of total VRAM without burning through a corporate enterprise budget Having that kind of massive memory pool means you can throw a ton of different large open weights models into rotation at the exact same time It completely simplifies your testing workflow by letting you switch environments instantly without running into local hardware bottlenecks

GITHUB JUST KILLED THE WORST PART OF VIBE CODING they shipped a free tool called Spec Kit and it already crossed 120,000 stars the fix is stupidly simple instead of tossing vague prompts at an agent and praying it doesn't wreck your project Spec Kit makes the AI write a full structured spec before it touches a single line of code it works through the problem first figures out what you want to build asks about the gaps lays out the project then it starts coding you get fewer insane bugs, cleaner output and results you can predict the flow looks like this: /constitution for your rules and standards /specify for what you want to build /clarify for the open questions before you start /plan for architecture and stack /tasks for the ordered work /implement to run it it plugs into Claude Code, Cursor, Copilot, Codex, Gemini CLI and 25+ other agents 120,000 stars, 10,000 forks, open source, shipped by GitHub itself learning to drive agents like this is most of what separates people getting hired as AI engineers from everyone still fighting their prompts

@CRS19 I just sent you a dm you can check your inbox

Definately possible

@CRS19 Well for starters, you do know that there is a permanent fix for it yes? You can follow me or leave me a dm and I’ll send you the details no problem

@WindexWix66725 @EndWokeness Hey... no offense, but it looks like your account is possibly shadow banned.

@EndWokeness If the Gop had balls after 9/11 there would be no Muslim gimmigrants in this country . We didn't mass import Japanese after they attacked us in 1941 .What makes these gimmigrants so special?

!!! NYC Department of Buildings rejects plan to save a 174-year-old church in Bushwick that burned via arson Demolition is set to begin in August Ahmed Tigani leads the agency

@cfs30425 @bennyjohnson @SenatorCollins Hey... just thought you should know, your account may be shadow banned.

@bennyjohnson @SenatorCollins is right. Its only 1 example on the crazy behavior that democrats have to stop. Schumer and company needs to grow up.

Senator Susan Collins Just CALLED OUT Chuck Schumer for Inciting Violence Against Supreme Court Justices 🔥 “For example, a Senator stood in front of the Supreme Court building and called out two members by name, saying that they had released the whirlwind and that they will pay the price.” “This overheated language, this completely inappropriate rhetoric against the Supreme Court endangers the lives of the justices and erodes public trust in our system of government”