Pierce Freeman
684
posts
@piercefreeman
San Francisco, CA
Joined January 2010
Pinned
Gonna figure out this damn X algo if it’s the last thing I do. Planning on posting more so I figure it's worth actually introducing myself. Here we go. 👇 I grew up in the Bay and watched this place evolve through every boom and bust it's had. My tech obsession started
starting an enterprise software company at 19 is like running away to join the circus except the circus had customers
Time is money and if you look around most SF offices these days, people are spending more than half of their time just waiting for llm outputs (even with multi-tasking) If this ends up costing $5k a month I can see a lot of companies paying. $5k << (300k/12) salaries
Replying to @OpenAI
Based on the Cerebras earnings call, the biggest challenge to getting everyone access to this speed are the boring things: -more data centers - more power - manufacturing more boxes Chip allocation is actually the lowest on the list since they manufacture on less constrained
I'm sorry but who in the Chrome/Gemini team thought it was a good idea to auto-add a gemini shortcut to my menu bar WITHOUT ASKING.
Rich and I got asked on our FAQ series about how an AI engine turns a question into an answer. without even talking about prefill or autoregression, here’s the usual pipeline: a question lands with a server and gets rewritten. most people under-scope their questions, or spell
