a famous electronics engineer says claude code rebuilt his multimeter's firmware from ONE prompt. plain c, under 24 hours, flashed and tested on the real meter which is kinda crazy that is #1 of 20 real things people are building with claude code right now. counted down in the video #2 anthropic's c compiler, written by 16 claude agents #3 a multiplayer heist game from 26 opus 5.5 agents overnight #5 a bridge test, opus 5.5 held about 130 lbs #8 a tv stripped of its bloatware #9 a crm from one prompt, six sub agents in two waves games, hardware, apps and the agents that run a business. for each one: who made it, how claude did it, and the builder's own demo #14 is a lawyer who cannot code LOL

Did ChatGPT just get nerfed? Has anyone else noticed some pretty significant changes to ChatGPT's usage limits? With ChatGPT seemingly combining regular ChatGPT usage with Work/Codex, I'm now seeing a weekly usage limit that I don't remember having to worry about before. Previously, when I was using Codex, my usage would reset every five hours. I could do quite a lot of development work, hit the limit occasionally, and just wait for it to reset. Now, I'm seeing a weekly usage bar, and the limits seem significantly more restrictive. For example, I had 10% of my weekly usage remaining today. I spent just 10 to 15 minutes using ChatGPT Work, and that entire 10% disappeared. That's a pretty massive difference from what I was used to. I use ChatGPT extensively for building browser extensions, developing apps, troubleshooting code, and working on various projects. If this is the new normal, it's going to seriously impact how much I can get done. I'm on the Plus plan, and until now, I've never really had an issue with weekly limits. I'm not sure whether this is a permanent change, a temporary adjustment, or something to do with how usage is now being calculated. Has anyone else noticed this? Are you burning through your weekly limits much faster than before, or am I missing something about how the new system works?

Hey @Beau Johnson , do you just use Gemini for these demonstrations or do you use them for personal business builds as well? I thought you primarily used Claude for that.

this video got me 100k views and claude code made the entire video end to end. there is no reason why you cant do this either opus 5.5 is your content creator the stack.. ----- 1/10 ----- I don't edit my long-form YouTube videos anymore. Every night at 4 AM, Claude Code wakes up, picks the topic, researches it, writes the script in my voice, makes the B-roll and graphics, and has a finished video waiting when I wake up. Here's the exact pipeline 🧵 ----- 2/10 ----- Step 1: the topic Two scouts run every morning: • a YouTube outlier scout (videos doing 5–10x their channel's normal views = proven demand) • an X scout for what's blowing up in AI right now It fills the next 7 days. My own ideas always come first. ----- 3/10 ----- Step 2: research + fact-check One agent pulls the primary sources: the company's posts, docs, repos and clips. A SEPARATE agent checks every claim and writes a "DO NOT SAY" list. No number goes in the video unless it's in the sources file. ----- 4/10 ----- Step 3: the script Written from a voice guide built from 10 of my real videos: my openers, my habits, my humor. Then a fresh agent grades it: facts, length, hook, one call to action. It fails? Rewrite. Fails 3 times? It stops and leaves me a note. ----- 5/10 ----- Step 4: the voice My ElevenLabs voice clone reads the script. Every chapter is checked to confirm the right model made it, because the old one sounds robotic. ----- 6/10 ----- Step 5: B-roll • A library of every clip I've used, tagged by company and product • Used clips rest 15 days before they come back • New footage from X via Grok + the X API, every post verified and credited Nothing repeats in a video. ----- 7/10 ----- Step 6: graphics 4–8 animated explainers, each with its own theme matched to the topic. Plus motion cards from @everestchris6's components, with a new look every video so they never feel the same. All built in code. No After Effects. ----- 8/10 ----- Step 7: assembly + QA A cue sheet lines every visual up with the exact words.

the #1 model on the leaderboard is one you CANNOT use yet lol gemini 4 carbon took the top spot on the artificial analysis index at 59, one point over opus 5.5. no public access same week openai cut the api price of its computer use model in half github copilot stopped making you pick a model. chatgpt pro and business get up to $500 a month in api credits anthropic filed to go public. google got an ai answer back from a tpu in orbit i ranked all 25 in the video: 5 headliners, 15 rapid fire, 5 trending repos, and whether you can actually use each one this week my call: carbon opens to the public before fable 5.5 ships