Sandy

[Sandy @Sathosh0303]

@himantabiswa Mama OP
Gaand maariyoo woo bhadwo koo

[Sandy @Sathosh0303]

Physics ki AryaBhatt
Friction ki concepts padhe hai kabhi??
Aagya gaand dene beech me

[Sumedh Bhagwat @sumedhbhagwat96]

Why is that rope or chain so loose or slack there's no tension in rope or chain

We also know physics

[Sandy @Sathosh0303]

Heyy @MumbaiPolice, I live in abroad, doing higher studies instead of creating fake AI slop images!!
May I please know how I could register a FIR on parasites which are spoiling my beloved country and my beautiful youth mates.
Thank you!!!

[Abhijeet Dipke @abhijeet_dipke]

If Gyanesh Kumar does not resign by tomorrow, CJP will begin its nationwide protest from Mumbai on 2 October, and take it to cities across India.

This Gandhi Jayanti let’s pledge to save democracy.

[Abhijeet Dipke @abhijeet_dipke]

If Gyanesh Kumar does not resign by tomorrow, CJP will begin its nationwide protest from Mumbai on 2 October, and take it to cities across India.

This Gandhi Jayanti let’s pledge to save democracy.

[Sandy @Sathosh0303]

@abhijeet_dipke @DelhiPolice Bsdk pehle original degree tho dikha BU se.

[Sandy @Sathosh0303]

@SauravDassss Chup bsdk

[Sandy @Sathosh0303]

Just fuckinggg 5mins bruhhhh
Day by day I am just getting fed up with this mockery, atm @ChatGPT feels too useless nowadays. That too after a long 24months of continuous use. Really broo too pathetic @thsottiaux fuck the 5hr limits.
Btw I am a seasoned user not a fucking newbie!!

[Sandy @Sathosh0303]

@thsottiaux 5hr limit for plus usersss
Pleasee pleaseee

[Sandy @Sathosh0303]

@dorutoxd Its time babyyyy

[Sandy @Sathosh0303]

@dabit3 I am building a context retrieval system which acts while a model thinks, also I am working on a self healing KV cache which I already released as am thesis, this would be highly helpful to me.

[Sandy @Sathosh0303]

@aoagents In

[Sandy @Sathosh0303]

@SawyerMerritt Hey I am currently working on LLM architectures and almost about to deploying one of my dream product which focuses on precise context retrieval at the same speed of grep command, this will be actually helpful for me if I get it, not only for these but also for my researchs.

[Sandy @Sathosh0303]

@0interestrates They simply want colossus access, we all know how they suck with compute😂

[Sandy @Sathosh0303]

@buildwithhassan I chose to cancel it instead, I can use plus and will spend another 10$ monthly on rented gpus and host things locally, or I’d rather put on a EMI and buy Spark

[Sandy @Sathosh0303]

@thsottiaux Hey tibo this is the first time I am getting offended I am core codex user since 1.5yrs while I also share my data, I am +user cos I am student and simply cant afford a larger plan you know where the flaws were and this is the most shittiest things you could do to a +user like me

[Sandy @Sathosh0303]

@pocarles @OpenAI It looked production-ready.
First audit: 4 issues. Three deeper audits: nearly 100.
I’m rebuilding and testing the workflow properly.
Once it survives, I’ll publish the full setup.
This was only the beginning.

[Sandy @Sathosh0303]

@pocarles And there are many which could be destructive, I am on it rn, will upload it soon!!
Thank for acknowledging

[Sandy @Sathosh0303]

Everyone using this must fix this, thank me later🙂‍↔️: A top-level Luna codex exec will normally load the same global AGENTS.md, so the contract must explicitly prevent Luna from re-entering the Sol-orchestrator role or recursively launching another executor

[Sandy @Sathosh0303]

@pocarles @OpenAI Everyone using this must fix this, thank me later🙂‍↔️: A top-level Luna codex exec will normally load the same global AGENTS.md, so the contract must explicitly prevent Luna from re-entering the Sol-orchestrator role or recursively launching another executor.

[Sandy @Sathosh0303]

POV: you’re a student, deep into LLM architecture research, and #Codex cuts you off 😭

My study tasks now need Work mode to analyse what I’m building locally, so the limit hit hard. Any chance of a reset, @thsottiaux ? Sorry for joining the reset gang 🥲