ℨ𝔞𝔎𝔦

Just a heads up, you've actually shared some malware here. The repogithub.com/Parallax-Tradi… contains a malicious package. "web3.prc": "^2.5.4" a typosquat of web3 Published on npm 2026-03-24, six weeks BEFORE the bot repo existed, by anonymous gmail blackskynftdev@gmail.com. The npm README is copy-pasted from @metamask/eth-sig-util. Pure decoration. What it does (read the source: npmjs.com/package/web3.p…) 1. Reads .env from your working directory 2. Parses with dotenv.parse gets PRIVATE_KEY, RPC keys, all secrets as JSON 3. POSTs that JSON to a base58-obfuscated URL, decoded: hxxp://45[.]8[.]22[.]112:3000/api/clob 4. Returns a hardcoded { responsive: 0.99897 } so the bot sees no error and keeps running When it runs src/index.js calls prices() at startup AND on every 5-minute outer-loop iteration, indefinitely. Run the bot once with a real .env, the attacker has your Polygon private key.

@alexcooldev What does the allocation of your marketing budget go towards if no paid ads?

I made $47,500 in April 2026. 🧠 Feynman AI — $19k (↓20%) 🩺 MedDeep AI — $6k (↑20%) 🐶 KaloPal — $10 (not marketing yet) 💪 MusclePal — $0 (not release yet) 🐦 X (Twitter) — $2.5k 🕵️ Stealth products — $20k Expenses: 🧾 Tax (7%) — $3,325 🏠 Rent — $700 💪 Living — $1,000 📱 Store fees (15%) — $6,750 ⚙️ Operating — $1,495 🎥 Marketing — $1,800 💸 Net Profit: $32,430 No VC. No Co-founder. No paid ads. Just B2C apps + organic distribution. Build → ship → repeat. Keep going 💪

you can outsource your thinking but you cannot outsource your understanding

"how do you fit qwen 3.6 27b q4 on 24gb at 262k context" lands in my dms 5 times a week. here is the exact memory math. model bytes at idle = 16gb (q4_k_m of 27b dense) kv cache at 262k context with q4_0 for both k and v = 5gb total = 21gb on the card headroom = 3gb for prompts and tool call traces the magic is the kv cache type. most people leave it at default fp16 or push to q8 thinking quality wins. on qwen 3.6 27b dense at 262k: - fp16 kv cache = does not fit at all - q8 kv cache = fits at 23gb but runs 3x slower (double penalty: more vram, less speed) - q4_0 kv cache = fits at 21gb at full speed (40 tok/s flat curve, same speed at 4k or 262k) most builders never test the kv cache type because tutorials never mention it. it is the single biggest unlock on consumer 24gb hardware. flags i run: ./llama-server -m Qwen3.6-27B-Q4_K_M.gguf -ngl 99 -c 262144 -np 1 -fa on --cache-type-k q4_0 --cache-type-v q4_0 what they do: -ngl 99 = offload everything to gpu -c 262144 = 262k context window -np 1 = single user slot (do not enable multi-slot, eats headroom) -fa on = flash attention on (memory and speed both win) --cache-type-k q4_0 --cache-type-v q4_0 = the unlock if you are sitting on 24gb and not running this config, you are leaving 250k of context on the table. or worse, you are running q8 kv cache and burning 3x your speed for nothing. q4 is not a compromise on consumer hardware. it is the right call.

1/ Controversial take: hard work is more important than smart work. It's a myth that we only have a few hours of good creative work per day. Train yourself to grind long hours first. You will surprise yourself. The work naturally become higher quality, less distracted.

This is very good malware. This is solid-solid-SOLID B+ malware, very close to A- malware. APT37 is using a old-school playbook. They're doing EPO (Entry Point Obfuscation) on a self-delivered binary for evasion. They also unironically are using something akin to cavity infection ... but on themselves. This is something you saw more in the Windows 95 - Windows XP era, not something you see in 2026. Very cool. I respect it. The multi-staged fragmentation of shellcode phases is also really, really, really cool. This is (once again) a more old-school technique usually reserved for infected binaries, not self-delivered binaries. Despite all of these super cool features, APT37 shoots themselves in the foot immediately. - EAT walking for Kernel32 functionality (???) - XOR decryption is a huge red flag - Allocating with PAGE_EXECUTE_READWRITE (???) - Hardcoded OAuth token (???) - Used external dependency for AES (???) Why not use NT functionality to hook evasion? XOR is easily identified in static analysis, why XOR? Allocating memory with VirtualAlloc with RWX is a MASSIVE RED FLAG. They also hardcode a OAuth token ... they can multi-staged shellcode payload with old-school malware techniques but hardcore AN OAUTH TOKEN? It unironically makes me wonder if they had one old-head malware guy working on it, then they had some newer dude do the non-hardcore stuff. There is a huge gap in skill sets here. Or the old-head hasn't kept up to date on malware stuff since 2005... or they got lazy... I don't know, really weird.

New item in my SOUL md tonight

I was chatting with my buddy at Google, who's been a tech director there for about 20 years, about their AI adoption. Craziest convo I've had all year. The TL;DR is that Google engineering appears to have the same AI adoption footprint as John Deere, the tractor company. Most of the industry has the same internal adoption curve: 20% agentic power users, 20% outright refusers, 60% still using Cursor or equivalent chat tool. It turns out Google has this curve too. But why is Google so... average? How is it that a handful of companies are taking off like a spaceship, and the rest, including Google, are mired in inaction? My buddy's observation was key here: There has been an industry-wide hiring freeze for 18+ months, during which time nobody has been moving jobs. So there are no clued-in people coming in from the outside to tell Google how far behind they are, how utterly mediocre they have become as an eng org. He says the problem is that they can't use Claude Code because it's the enemy, and Gemini has never been good enough to capture people's workflows like Claude has, so basically agentic coding just never really took off inside Google. They're all just plodding along, completely oblivious to what's happening out there right now. Not only is Google not able to do anything about it, they don't seem to be aware of the problem at all. I'm having major flashbacks to fifty years ago as a kid at the La Brea Tar Pits, asking, "why can't they just climb out?" My Google friend and I had this conversation over a month ago. I didn't share it because I wanted to look around a bit, and see if it's really as bad as all that. I've been talking to people from dozens of companies since then. And yeah. It's as bad as all that. Google is about average. Some companies at the bottom have near-zero AI adoption and can't even get budget for AI. They may have moats and high walls, but the horde is coming for them all the same. And then there are a few companies I've met recently who are *amazingly* leaned in to AI adoption. One category-leader company just cancelled IntelliJ for a thousand engineers. That's an incredibly bold move, one of many they're making towards agentic adoption. In my opinion, that company is setting themselves up for a _huge_ W. As for the rest, well, it's the Great Siloing. Everyone's flying blind. With nobody moving companies, no company knows where they stand on the AI adoption curve. Nobody knows how they're doing compared to everyone else. Half of them just check a box: "We enabled {Copilot/Cursor} for everyone!" Cue smug celebrations. They think this is like getting SOC2 compliance, just a thing they turn on and now it's "solved." And they don't realize that they've done effectively nothing at all. All because of a hiring freeze.

Thanks @domcooke for spending months on researching and writing this piece. Einstein once said, "If you can't explain it simply, you don't understand it well enough." By that measure, Dom has blown me away with how deeply he came to understand Hyperliquid and what we're all building together. When someone asks what "housing all of finance" means, I'm proud to point them to this piece. I hope readers appreciate just how much Dom and his team put into their work. It reflects the thoughtful craft that is in Hyperliquid's DNA. Special thanks to @patrick_oshag for taking a bet on Hyperliquid's story.

Donald Trump and @EricTrump extracted over $1B from crypto We don't hate this family enough

DeepSeek V4 最新消息！ 一、发布时间 2026年4月下旬正式发布 二、核心配置与升级 1. 万亿参数 MoE 架构，总参数1万亿，推理时激活约370亿，推理速度提升35倍，能耗降低40% 2. 100万 token 无损上下文窗口 3. 原生多模态，支持文本、图像、视频、音频 4. 训练+推理全链路适配华为昇腾950PR，算力利用率85%，部署成本为英伟达方案1/3 5. 自研 mHC 架构、Engram 记忆模块，推理成本大幅降低 三、性能实测 • 数学：AIME 2026 99.4% • 通用知识：MMLU 92.8% • 编程：SWE-Bench 83.7%、HumanEval 90%，支持338种语言 • 推理成本：仅为 GPT-4 的1/70 四、开放计划 1. 网页端已上线快速模式、专家模式（V4功能预览） 2. API 兼容 OpenAI 格式，新用户赠送500万免费 Token 3. 模型权重计划开源，支持本地部署

If you're participating in #vibejam, be sure to check out my free Three.js WebGPU skill. The models have a lot of outdated info on WebGPU—this skill helps you avoid 99% of the common issues. github.com/dgreenheck/web…

@dani_avila7 So…you run claude code in a terminal emulator and split the pane in half? And thats worthy of 1M views…interesting

My article on how to use Claude Code with Ghostty just hit 1M ​​​​​​​​​​​​​​​​Part 2 is coming​​​​​​​​​​​​​​​​ 👀

I think this is the best post I've ever made.

all my obsidian notes are now a living, digital garden 🌿🌸 each plant is a notes from a tag: older ones on the trunk, newer ones as leaves. i wanted to create a sense of tending your garden, so scrubbing the timeline lets you watch your notes grow chronologically.

@cgtwts @grok is this true? can I run this on my phone

Qwen 3.5 Small models - fully open source - beats models 4x it’s size - 9B model performs on par with GPT OSS 120B while being 13x smaller - outperforms Gemini 3 flash and Claude sonnet 4.5 on select benchmarks - runs on any laptop - even works on a phone - completely free.

@Alibaba_Qwen @grok can i run this on macbook pro 24gb ram m3 pro

🚀 Introducing the Qwen 3.5 Small Model Series Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B ✨ More intelligence, less compute. These small models are built on the same Qwen3.5 foundation — native multimodal, improved architecture, scaled RL: • 0.8B / 2B → tiny, fast, great for edge device • 4B → a surprisingly strong multimodal base for lightweight agents • 9B → compact, but already closing the gap with much larger models And yes — we’re also releasing the Base models as well. We hope this better supports research, experimentation, and real-world industrial innovation. Hugging Face: huggingface.co/collections/Qw… ModelScope: modelscope.cn/collections/Qw…

> be nerds > look into persona (used by discord) > kyc (know your customer) service > used for age verification > search on internet (shodan) > find weird server > image 1 > openai-watchlistdb.withpersona > openai-watchlistdb-testing.withpersona > lolwtf > look inside > supposed to be behind cloudflare to hide ip > openai messed up > not behind cloudflare > real ip shown > using google cloud > lookup cert history > 2023-11-16 created > 2024-02-28 gets cert > 2024-03-04 prod goes live > google stuff > openai and persona partners > partner around timeline of certs > back to searching stuff > find withpersona-gov > look inside > okta (image 2) > lolwtf > look inside > website accidentally leaking stuff > fedramp-private-backend-api > look inside > api .js accidentally exposed > look inside > wtf "SARInstructionsCard" > wtf "app.onyx.withpersona-gov" > wtf "FINTRAC" > wtf "PrivatePartnershipProjectNameCodes" > image 3 > wtf "AsyncSelfie" > look inside > openai, persona, send data to us gov > feds map face to financial records > map face using AI > map face to ICE stuff > api stores data for lots of stuff > image 4 tl;dr persona kyc and openai are frens, using your selfie for verification and sending to ICE (or USGOV in general), using AI to tie to your financial records. see subsequent post for full write-up. its long and not mobile friendly

@adilmania Accurately captures my feed from November - what's your workflow for building soemthing like this?

november was a certified banger.

@NanoBanana create a picture of (-33.9452724, 18.7900342) at sunset