The infrastructure layer (SpaceX) captures value first before the application layer (Tesla), but the application layer has a more durable moat
Love to hear that “B2B is our fastest growing category with companies increasingly using Grok Bot”
Grok Bot already has 800k paid weekly users—doubling month over month
The cursor acquisition was massive!
What is the journal smoking?
Unveiling the Roadster with out-of-this-world technology reiterated that Musk has given up designing cars for regular people to drive. on.wsj.com/3UvugrS
Just a reminder that SpaceX is approaching its IPO price ~$130 🤤
This Meta-Apple post gave me a lot of ideas. Sadly, my Macbook Pro is old (2021), so I can't run any of the cool open models.
I realized the tiny open models could work.
Ended up with Qwen 3.5 4B because it was the best at the benchmarks Claude ran.
And man, it feels awesome to run compute locally and own something in the model stack (even if it's just fine tuning a tiny model).
I'm just frustrated Alibaba made Qwen, not Silicon Valley.
Man, I'd like a tiny american model!
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
