Most of the people I know would rank predictability is #1, price #2 and ability to do returns #3 (as a data point: I don't even remember last time I did a return, was years ago). Really depends on who your connections are and their specific needs I reckon.

Not the smartest engineer ever, that use case perfectly align with using a Chinese model! And anyway I have no idea how he was hoping not to be caught in the case the device would be discovered.

I don’t agree with the decision, but I don’t think courts should be able to easily overturn the federal government’s findings when it comes to national security in a specific decision like this.

My toddler gets a book a month in the mail from Dolly's Imagination Library and we keep reading some almost every day. She's made an impact in so many families <3

Just tried and although it’s not that polished or fast imo it’s very cool.

GLM 5.2 totally free even with a free Vercel account.

Make business sense too - fx is the entry point to bring more user to Vercel AI.

Not sure as OpenAI models (Sol, Luna,..) are also discounted on the Vercel AI Gateway rn. My bet is on OpenAI trying to drive more enterprise customers to their models through API.

I’m impressed by tinygrad. Full-featured deep-learning framework with ZERO dependencies, fully hackable, and orders of magnitude fewer LoC than PyTorch.

Want to run Qwen3.8 27B? Download the quantized model from HF, git clone tinygrad, then run `DEV=METAL python3 -m tinygrad.llm --max_context 8192 -m `. Read the code and experiment - takes 1 minute.

Compare that with PyTorch (great, but almost impossible to comprehend and modify, IMO) or Mojo/MAX (complex and not even fully open source, not to mention that the Modular folks aren’t particularly nice). Tinygrad is a breath of fresh air.

"Zero dependencies" except Python, which is a pretty huge dependency.

The real target for Anthropic, OpenAI and tons of others are large enterprises and workflows. I think they will do all right if they execute well.