Viewing profile — tomr75

hi nsa

people who can't afford Claude max 200 are using qwen 3.6 27b for local coding assistance already

pretty sure Nvidia GPU is better bang for buck because of usable inference speed..

why V3 over V2 (assuming English only)?

use USD denominated stablecoins on a permissionless blockchain (any layer 2 ethereum network like arbritrum) or a ZK L2 ...

why can't it just use tool calling for RNG?

just reduce the services you provide Europeans, let the Europeans vote out these people

you can buy these sensors off aliexpress cheap

With qwen code?

flip-side is one person can add a billion dollars of value

it depends on how many people you expect to date in your lifetime though these days with dating apps can prob date way more than 18..

why use whisper over parakeet? how will you monetise?

presumably their apply model is run on their servers I wonder how hard it would be to build a local apply model/surely that would be faster on a macbook

expressive: https://github.com/resemble-ai/chatterbox dialogue like notebooklm: https://github.com/nari-labs/dia

how does it have access to DOM? are you using it with cursor/browser MCP?

I’m getting 56 with mlx and lmstudio. How 76?

someone else will do it if they don't

also RAG implementation, and the apply code feature of cursor..

have you tried cursor?

try cursor

can you use TTS models?

Any recommendations?

OVH did the same thing to me...

How has it worked well? Europe is having issues with productivity — too expensive to live there AND higher paying jobs in the US. Eg People have to leave Norway to start businesses…

which size are you using? I don't see why you would use it over claude and 4o-mini with cursor unless you are working on a top secret repo