@thestanduppod@DefNotZap Qwen 3.5 32B is sonnet 4.5 level intelligence but can be run locally at decent token speed (given enough ram). There are many benefits to being able to run performant models on local hardware e.g file system isolation, cost savings (250m tokens IO break even), data privacy, etc.
@Param_eth just like everyone should get a free personal trainer and private chef… if you can’t afford $20 a month rn for access to the smartest tools on earth you’re ngmi
Two types of people right now:
Group 1:
- “AI will take my job”
- “Market is dead, nothing works”
- “Everything is saturated”
- “Only insiders make money”
- “It’s too late to start now”
- “I missed the opportunity”
Group 2:
- learning tools nobody taught them
- building things nobody asked for
- posting when nobody is watching
- stacking skills while others panic
- becoming dangerous while others complain
- quietly positioning for the next 10 years
AI won’t decide your future. Your group will.
Which group are you in?
agent now triages every incoming issue from slack and sentry, feature request, gh issues, etc. in real time → re-ranks them strictly by my priorities + keeps track of status of all as commits/prs land on branches + sends me a message every morning with things i need to do that day. adding an auto-fix bot for small issues soon
timing wasn’t the issue. ubi would make this country insufferable. ppl already complain endlessly abt working & most choose to do nothing about it. great in theory if people start ventures but hard to see it working in practice. gov already subsidizes complacency every day - why make it worse??
@kale_abe 🎯 🎯 claude code is literally getting better by the hour with every new tool/tweak. classic Jevons paradox - efficiency removing friction -> usage/impact explodes.
the only people saying this won’t happen are those who have something to lose and/or are unwilling to learn
Introducing Perplexity Computer.
Computer unifies every current AI capability into one system.
It can research, design, code, deploy, and manage any project end-to-end.
@joanrod_ai@QuiverAI@a16z rarely do i find logo generators any good - this is actually a game changer. I’ve only been playing around w it for 5 mins and it’s seriously impressive
@Alibaba_Qwen this is the first model i’ve seen to make mac mini worth it. Break even cost at abt 250m tokens IO compared to sonnet 4.5 + not sending anthropic data + can run 24/7 doing many things at once