web16z

@NanouuSymeon C and JavaScript

As A Developer, what is your favorite programming language?

@ChShersh True SWE doesn’t need an AI to produce bad code 😄

Garry Tan produces AI slop and gets dunked on X for poor code quality. I manually and carefully handcraft C++ code and get dunked on X for poor code quality. But we’re not the same.

@GithubProjects Hi there 👋 github.com/ultroned

-▬▬.◙.▬▬‐ ▂▄▄▓▄▄▂ ◢◤ █▀▀████▄▄▄◢◤ █▄ █ █▄ ███▀▀▀▀▀▀╬ ◥█████◤ ══╩══╩══ ╬═╬ ╬═╬ Just dropped down to say ╬═╬ Share Your GitHub Profile ╬═╬ And Let's Connect ╬═╬ ╬═╬ ☻/ ╬═╬/▌ ╬═╬/ \

I found the exact session file for that chat in the app's local data and described to Claude in a new chat what had happened. It quickly figured it out and restored the chat. I can share the details if needed. This is something your team should look into @bcherny

UPD: the chat that caused it was very important to me, but the Claude app refused to work and froze every time I tried to open it. Reinstallation didn't help.

1.76TB ☠️ I beg your pardon @claudeai @bcherny @trq212

@devdiary0x @om_patel5 It’s called KISS and it’s opposite to bad in most cases

@om_patel5 Just hope it doesnt write caveman code "code too tuff, me write easy code"

I taught Claude to talk like a caveman to use 75% less tokens. normal claude: ~180 tokens for a web search task caveman claude: ~45 tokens for the same task "I executed the web search tool" = 8 tokens caveman version: "Tool work" = 2 tokens every single grunt swap saves 6-10 tokens. across a FULL task that's 50-100 tokens saved why does it work? caveman claude doesn't explain itself. it does its task first. gives the result. then stops. no "I'd be happy to help you with that." no "Let me search the web for you" no more unnecessary filler words "result. done. me stop." 50-75% burn reduction with usage limits getting tighter every week this might be the most practical hack out there right now

@RbdJellyfish Congrats man. In an ocean full of slop, good to see people still building real stuff.

✨ A Little Perspective is out now! ✨ I put my all into this strange, experimental game for the past 2.5 years, and I'm so excited to finally let you all experience it for yourselves. Head over to the Steam store to buy it now (10% off for the first week).

@gdechichi @ThePrimeagen Yes, Gabriel 💯 Make it happen Prime

@ThePrimeagen I'm building a game engine in C x.com/gdechichi/stat…

Hey, you got a cool project that you are building? Link it I want to yap about cool projects

Pretty accurate 👌

@masterbugstudio @PixelFrogStudio Looks good! But is it handcrafted, AI or hybrid?

My first demo of a game I’ve been passionately working on. I don't have any follower but still hope someone like it! then maybe I’ll keep posting updates here. Big thanks to @PixelFrogStudio for the Tiny Sword assets. #gamedev #pixelart #indiegame

@bunjavascript Bold move, and it's not a compliment

In the next version of Bun Bun gets 4 MB smaller on Linux x64 because we deleted CMake

@CescQuintero @ChShersh Looks like the full enum behind dating app replies lol

@ChShersh Holup

I've just seen the 2nd-worst enum in my life

@fuolpit @heynavtoor

@heynavtoor 🫡 github.com/angristan/nocl…

🚨 BREAKING: Someone just rebuilt the entire AI assistant stack in Zig. It's called NullClaw. The binary is 678 KB. It uses ~1 MB of RAM. It boots in under 2 milliseconds. No runtime. No VM. No framework. No garbage collector. Just raw Zig. Here's why this is absurd: → OpenClaw needs a $599 Mac Mini and 1 GB+ RAM → NanoBot needs 100 MB+ RAM and Python → PicoClaw needs 10 MB RAM and Go NullClaw runs on a $5 board with 1 MB of RAM. Same functionality. 0.1% of the resources. Here's what's packed into that 678 KB: → 22+ AI providers (OpenAI, Anthropic, Ollama, DeepSeek, Groq, etc.) → 13 chat channels (Telegram, Discord, Slack, WhatsApp, iMessage, IRC) → 18+ built-in tools → Hybrid vector + keyword memory search → Multi-layer sandboxing (Landlock, Firejail, Docker) → Hardware peripheral support (Arduino, Raspberry Pi, STM32) → MCP, subagents, streaming, voice, the full stack Here's the wildest part: Every subsystem is a vtable interface. Swap any provider, channel, tool, memory backend, or runtime with a config change. Zero code changes. It even encrypts your API keys with ChaCha20-Poly1305 by default. 2,738 tests. ~45,000 lines of Zig. Zero dependencies beyond libc. 100% Open Source. MIT License.

@TheGeorgePu It seems the best way for AI companies to make money is through insider trading and shorting the market on hype news rather than, you know, building a sustainable business model lol

Anthropic published a blog post one hour ago. Cybersecurity stocks have lost $10B since. CrowdStrike -6.5%. Cloudflare -6%. Okta -5.7%. One blog post. One hour. $10B gone.

@gdechichi Great read, I completely agree on that. I’m a little bit old-fashioned (or rigid lol) so I don’t use AI assistants to write code or solve issues, unless I’m in rush. But man, I cannot appreciate more having it to learn new things or just explain how/why certain things work.