thetopham

[thetopham @thetopham]

@notkevinzhang @typesafeai @CompleteSkeptic

[thetopham @thetopham]

@justKDeng

[thetopham @thetopham]

@justKDeng financial/market decisions/day-trading require low latency https://t.co/vtSoxrwemx [email protected]

[thetopham @thetopham]

@notkevinzhang financial/market decisions/day-trading https://t.co/vtSoxrwemx [email protected]

[thetopham @thetopham]

@mark_k @Sentdex @perrymetzger @MindsAI_Jack @alexwg inv

[thetopham @thetopham]

@StefanoGPT Windows app package version 26.901.6511.0 , increased usage compared to when it first came out. pro x5 gpt 6 astra ultra. first day, 44% to 1% took a full afternoon, 8-10 hours? developing a phone app. then reset yesterday, now its at 100% to 46% in 2-3 hours creating unity game

[thetopham @thetopham]

@icyveins arg's are the best

[thetopham @thetopham]

@34frosty how is it 334 already?

[thetopham @thetopham]

@sama @OpenAIDevs @gdb @karpathy @thsottiaux update: i don't want to export/migrate anymore lol. gpt is the database now= memory, projects, chats, gmail, health, finances

give us OAuth/API access to our own user chat context so apps, wearables & agents can use it

https://t.co/46FJJgJO5a

@sama @OpenAIDevs @gdb @thsottiaux

[thetopham @thetopham]

@sama chatgpt should support this natively:

export years of chatgpt history → use a simple llm-wiki skill → turn conversations into .md “exocortex brain” → agent browses/update in obsidian + github.

process:
https://t.co/tF6QZXsnIL

@OpenAIDevs @gdb @karpathy @thsottiaux

[thetopham @thetopham]

@trevorwood222 this increases moe speeds with gpu + ram

https://t.co/t1gCUv9FgV

[Shuo Yang @Andy_ShuoYang]

Your gaming PC can now serve frontier models at interactive speed using official checkpoints without extreme quantization!

Qwen3.6 35B → 8GB RTX 4060 laptop @ 39 tok/s

DeepSeek-V4-Flash 284B → RTX 5090 desktop @ 22-25 tok/s

GLM-5.2 753B → RTX PRO 6000 workstation @ 15 tok/s

Run your claude code or codex now with frontier model for $0

Meet FreeToken 🧵

[thetopham @thetopham]

@Andy_ShuoYang 3060 12gbvram, 32gb ram
3060 12gbvram, x2 dual gpu, 32gb ram
amd r9700 32gbvram, 32 gb ram
linux headless server 64gb ram (integrated gpu lol)
linux headless server 32gb ram (integrated gpu lol)

[thetopham @thetopham]

@NVIDIAGeForceFR @ASUS_ROG_FR #RTXPowersPlay

[thetopham @thetopham]

@NVIDIAGeForceDE @DigitalStormPC #RTXPowersPlay

[thetopham @thetopham]

@NVIDIAGeForceDE #RTXPowersPlay

[thetopham @thetopham]

@NVIDIAGeForceFR @DigitalStormPC #RTXPowersPlay

[thetopham @thetopham]

@NVIDIAGeForce @DigitalStormPC #RTXPowersPlay

[thetopham @thetopham]

@NVIDIAGeForceUK #RTXPowersPlay

[thetopham @thetopham]

@NVIDIAGeForce #RTXPowersPlay

[thetopham @thetopham]

@Keemotionx Gpt 6 before gta 6

[thetopham @thetopham]

@horizon_trade_x https://t.co/UFP1To4VpC