@TheBlackSheep84 “think critically” on bio and yet you have no idea what being on a different timeline means or why it happened and how it happened, with no evidence for it other than our memories being different. we simply misremembered and got influenced by rehashes and remakes. blocked
We're giving away a CONTROL Resonant custom wrapped NVIDIA GeForce RTX 5080 to celebrate its official release with #RTXON.
To enter:
🟢 Share this post
🟢 Comment #RTXON
T&Cs: https://t.co/Mbf7w3j1q3
@nagarete_cyber@0xSero the point of this post is to show how open weight local LLMs are catching up. obviously you cannot feasibly run local models that are at subscription level at all, but 6-10 months from now we will have opus 5.5 capabilities locally. progress is progress.
@TokenGremlin i understand it. im on a $200 sub and ive estimated that ive used around $12k-$25k worth based on tokens and models ive used. so it’s a steal imo
@bridgebench why +-10% is variance? is that arbitrarily picked? surely there is a better way to pick variance value for model nerfage. should be informed from past model nerf data. also, if a company can reliably nerf their models by 9% on your scale, you will always think they’re normal.
@thsottiaux not to sound ungrateful but you did say to burn our tokens and not that we’d get a banked reset, so why wasn’t it a regular reset? we wouldn’t have had to burn all our tokens if you would have mentioned it earlier that it was a banked reset. also i still havent received it.