tok/sec is really dragging here
COMPUTING
-
Ideal Phone Experience: Simplicity vs Features
By
–
what do you think the ideal phone experience would look like? would it be something simple and streamlined or something complicated with every possible feature even if it's hard to understand? is there any killer feature you want in it?
-
AI-Powered Browser as Future Operating System Platform
By
–
I think the browser with AI becomes the new OS. Building an actual OS requires too much work, for now. Maybe in a few years it can evolve into an actual OS when AI can just code the entire thing.
-
Building AI Agents That Understand Personal Computing Needs
By
–
Great podcast episode on how we're building with AI and agents that get things done. Our computers are getting more personal than ever with AI, they can now understand our needs, specific apps we use, and are able to get things done. Really enjoyed our conversation, thanks for
-
FlashAttention less useful with MLP GEMM latency dominance
By
–
Maybe FlashAttention is not that useful when you have MLP GEMMs that eat so much latency? Interesting graph in the latest blog post from @gpus_go_brrr!
-
One Code Line Reduces PyTorch Downloads by 400MB
By
–
one line of code just reduced pytorch download sizes by 400MB . for literally everyone lol so much low hanging fruit when you pay this much attention to detail
-
Comet Web Browser Launch for Modern Internet
By
–
Comet is here.
— Perplexity (@perplexity_ai) 9 juillet 2025
A web browser built for today’s internet.pic.twitter.com/cFPeghl2YMComet is here. A web browser built for today’s internet.
-
India Launches E-Passport with Automated Border Control Gates
By
–
Finally got latest Indian E-Passport. E-Gates here we come.
-
Local coding assistants protect privacy without external servers
By
–
Most coding assistants send your code to external servers, even for simple tasks like reading files or running shell commands. That’s a problem if you care about privacy, work in secure environments, or just want full control.
-
Cloud limitations: why small local models matter more
By
–
Cloud models just can't offer some features by design: privacy, low-latency, offline connectivity, free generation, etc. Not every application requires 2T parameters. Probably most don't.