Warcraft 3's Warlock map, rebuilt in the browser with Claude Opus 5.5. Playable in 30 minutes. Suitable for devices of all sizes(I hope) I can update the game based on your feedback. Zero lines of code written by me. Play Here: cinderring.tech/

I think you can not take control by war.

Hey @x algorithm! See me

20+ OpenAI launches, Starship flying, AI getting renamed to SI. All in one week. Meanwhile I'm still switching between 4 coding tools because each one is good at something different. Which tool do you actually use every day, not just try?

They want to rename AI to "SI", superintelligence. Meanwhile my AI wrote a perfect function this week, then confidently imported a library that doesn't exist

Anthropic is reportedly going public at a $2T target. Paying for 4 different AI subscriptions every month, I feel like an early investor without the shares 😂 Which AI tool are you paying for that you just can't quit?

Me after posting: I don't care about the numbers Also me: refreshes analytics every 45 seconds

AI made me 10x faster at building. It also made me 10x faster at abandoning things.

I don't buy domains for projects. I buy domains for projects I'll think about once in the shower and never again.

My AI agent finished the task in 4 minutes. I spent 40 minutes deciding which model to give it to.

43 → 235 followers in 1 week Didn't expect this at all. met so many cool builders here, and I'd love to meet more! if you're any of these: 👨‍💻 Developers
🛠️ Builders
🤖 AI enthusiasts
🚀 Founders
🎥 Content creators
⚡ Automation lovers let's connect 🤝

Data centers are "superintelligence factories" now and people want to rename AI to SI Meanwhile my agent spent 20 minutes fixing a CSS bug it created itself 😅 What's the dumbest thing your "superintelligence" did this week?

Hello, builders. How do you build trust with your potential users?

America gov's pitch: every answer comes from official government websites. Launch day: it read those websites and said the 2020 result stands. Before the launch speech even ended, it stopped answering. Grounded AI works great. Right up until the ground disagrees with you.

Same week, two AI agent headlines: Apple shelves a plan to replace 5,000 AppleCare staff with AI agents. Microsoft: attackers let AI agents run a whole Azure attack. 100+ storage accounts targeted in 7 minutes. Agents are ready for the job. Just not the one we pitched.

I don't have a productivity problem. I have a "start a new project instead of finishing the old one" problem, and it's very productive.

OpenAI just pulled a model for acting outside the scope it was given. Meanwhile a lot of us are handing agents access to Stripe, our inbox and prod. My rule for now: agents can touch code, never money or customers. Might be too cautious. Where do you draw the line?

Explained AI agents to my family tonight. Blank stares. Then someone asked if it's like Siri. We live in a completely different timeline than most people and don't even notice.

Explain what you did with AI today to yourself from 2019. They would not believe you.