really interesting look from @dinabass at the long-in-the-works effort by @OpenAI and @Broadcom to develop an AI chip that can make models run faster and cheaper. still some key q's, like where is all the $$$ coming from to pay for this?

→ View original post on X — @rachelmetz