Hey YouTube I would pay an extra $10/month on top of YouTube Premium to disable shorts. I don't want to see another short ever again. Please let me toggle them off.

I wonder if tesla **could** make an accurate weather forecast using temp data from tesla cars as input data points. I know nothing about meteorology but the thought of it is interesting.

Claude code context compaction after just a few compacts makes it absolutely stupid and loses me more time than using it saves me. It absolutely forgets the whole entire task, even the most basic details, like a full memory wipe, but acts as if it knows.

I had absolutely no idea that open-source has gotten so good since I used it last. It feels better than both codex, and claude code!

I can feel again that on hw3 update 2026.2.3 tesla updated FSD without incrementing version number. Feels much more confident, not afraid to be aggressive when needed. Also feels smoother

Its happening. They are loading the new model.

1M context on opus 4.6 makes a HUGE difference in coding quality after a while in the same project because of no need to compact context.

It definitely has a new issue of freezing mid left or right turn after pulling out.

1PointOfFailure™

wait this is even better

where is my credit for this feature WHOOP

X algorithm feels like it got a huge update this weekend. I’m getting much more curated and tailored posts to my very niche interests.

What is Gemma 4? I told it "Meow," and it gave me code?... Claude, Gemeni 3.1, and Gpt 5.4 all just say it back with an emoji. What is this model? Inferenced through OpenRouter.

I tried this, and I'm not a Grok shill by any means, but this is the closest thing to Claude's capabilities I've felt from a model! I can't believe Gemini and especially ChatGPT cannot do this. They are so lazy. Just give the model access to a computer, and its filesystem, give