Anson

Switching to Chinese semantic compression, cus English is inefficient for AI. 畫蛇添足 = "Do not add unnecessary steps, explanations, or features" 對症下藥 = "apply the right fix to the exact problem" 條理分明 = "organise the response into clearly structured sections with a logical flow, so each part builds on the previous one and is easy to parse" ~4 tokens vs ~15+ in English. Same idea. Half the cost. 0% tokens left. Time to get efficient.

@alexdolbun 🤝👨‍🌾

@ansonox 👨‍🌾🤝

@UnslothAI @Zai_org @0xSero can this be further optimised?

GLM-5.1 can now be run locally!🔥 GLM-5.1 is a new open model for SOTA agentic coding & chat. We shrank the 744B model from 1.65TB to 220GB (-86%) via Dynamic 2-bit. Runs on a 256GB Mac or RAM/VRAM setups. Guide: unsloth.ai/docs/models/gl… GGUF: huggingface.co/unsloth/GLM-5.…

@austingriffith one thing I disliked about hackathons is that many people just focus on the prizes instead of building something beyond that.. nothing wrong with it, just feels short-sighted..

📑 Stop trying to pad your resume and get hired... 👨‍🔧 Go build something people actually use!! 🎟️ No more hackathon bingo to pay for your plane tix. ✈️ The upside is massive, think: private jets!!!

This is a little beast! huggingface.co/Jackrong/Qwopu…

@jakobjelling x.com/ansonox/status…

It's Monday. Time to promote your product. What are you working on?

@halenhargreevs x.com/ansonox/status…

You're old, but are you this old?

Launching Giggle.local A search engine with no internet. ✅ a farm ✅ solar panels ✅ local search Just open it locally, search like usual, and click any result to generate the website on the spot. Truly the missing piece of the tech bro dream.

@AlexEngineerAI Tried to actually use it to code in both codex and @opencode , it just get stuck in tool calls.

ngl. qwen 3.6 plus and MiniMax M2.5 are really good. and they are free now

@felixhhaas Lovable

Ultimate AI Founder Playbook 🔥 Over the past few years I've invested in 30+ startups and worked inside one of the fastest growing AI companies in the world. I distilled everything I've learned about building something people love into one playbook. Here's what you'll get: 👉 My framework for building products people love 👉 What makes a great founder in the age of AI 👉 How to go from idea to working product in hours 👉 How to turn users into your best growth channel Super excited to be giving a lecture about this tomorrow at our Stockholm HQ. Come join us if you're in town! Can't make it in person? Comment "Lovable" and I’ll send you a link to the playbook! LFG 🚀

We need one for the local folks @lmstudio

@OpenAI @thsottiaux @Dimillian wait does using 5.3-codex-spark also use the weekly usage limit? @Dimillian

my usage just went down to 0% even though i haven't used any today? pretty sure i had at least 2% (lol) this morning! @OpenAI @thsottiaux @Dimillian

a great Sunday read! the guy, Jackrong who created qwen3.5-27b-claude-4.6-opus-reasoning-distilled wrote a whole pdf teaching you how to fine-tune your LLM. github.com/R6410418/Jackr…

@thekitze any good?

FYI if ur using paperclip i just found out that it takes 40gb of disk space because of backups 🤦‍♂️

@thekitze LMAO you mofo 🤣

get free chatgpt pro with this one simple trick (I hate this pattern so much)

@pbteja1998 @openclaw Me the other day 🤣x.com/ansonox/status…

Tried GPT 5.3 Codex with @openclaw, and it's so bad... and it also is giving me the famous... "If you want, I'll now...." thing at the end 😵 Not sure what's the best model to use with @openclaw now... I started a new session and asked that same question to Sonnet, it's response is easily 10x better.

@philkellr spending 5x the time fixing my agent + burning the tokens that you paid for to fix it.

I'm tired of OpenClaw Every 2-3 days I have a major moment which I show to my friends: "look what AI agents can do", but then the other 90% is pure frustration and me cursing at my own AI agent for which I spent hours choosing a beautiful name and profile picture. I was fun at start: I added telegram, added voice input, then adding skills even with just voice prompting was a bliss. Then dementia hit. Facts from 48+ hours ago were forgotten. I installed a 3 level memory system. It felt like a huge hack, it barely works, I encounter bugs every day which I'm fixing. Then it breaks with every update. Not all of it, but little things. The WhatsApp integration is just insanity. After putting in every markdown file that it shouldn't reply to my friends (IN CAPITAL LETTERS) it happily started to chat with my wife and my goddaughter. And from today even claude subscription stops working. I feel like a failure! I see all the success stories left and right, YT vids and blog posts "I got it to work and here's what the AI agent does for me", and for me I'm still spending 5x the time fixing my agent than just doing the stuff by hand 🤷