peter pork

porkz
·

AI & ML interests

AI & ML interests

Recent Activity

Organizations

Open-Source AI Meetup's profile picture University of Utah's profile picture KAISAR's profile picture open/ acc's profile picture

porkz's activity

replied to fuzzy-mittenz's post 5 days ago
view reply

never tried "do better", but i will for sure.

I usually think like i'm an legion of zombies and it's always "we", curiously DeepSeek (even new V3), Gemini and Reka accept that easier, and instead of "the user wants.. but wait user is making fun, o h wait user is saying other things (when they perform a search and blame me for bad results)" , when they assume that "we", no more time wasted pretending to be dramatically alarmed with what I've said.

Now I'm using "we" all the time with any model. And what we like about it is that since the AI and I are one, they stop with those lame "you's" and in fact do something for us

replied to fuzzy-mittenz's post 6 days ago
view reply

i agree with this.

they are overrated.

reacted to fuzzy-mittenz's post with 👍 6 days ago
view post
Post
709
So frustrated with "Reasoning" Models.
Sure, introducing RAG into the mix, or giving it an interpreter to math with helps, but never as much as a model that has good instructions.

Even if it's just to repeat the information before answering, a normal model will usually out "Think" it's reasoning counterpart.

Not sure if it's my frustrations but the best answers I've received (from a reasoner), so far, are from the simple instructions to, "Do better!"

Figured I would share the special sauce.

Using 10-100x Compute just to heat the office can't be environmentally friendly, and It still has no Idea where my keys are.
  • 4 replies
·