Arison Daniel

You'll come to know me one way or another

[Arison Daniel @xzeronelysium]

We're not talking about it enough but current SuperGrok usage is very ass 😔

[Arison Daniel @xzeronelysium]

It's been a long time but now I remember why I have a love and hate relationship with DOCKER 🤦‍♂️

[Arison Daniel @xzeronelysium]

@0xSero spacexai?

[Arison Daniel @xzeronelysium]

@jackfriks Termux is better imo and you can really copy and paste and all that stuff

[Arison Daniel @xzeronelysium]

@mkratsios47 Dumb fuck lol

[Arison Daniel @xzeronelysium]

@levelsio @DavidOndrej1 answer the question bruv 👽

[Arison Daniel @xzeronelysium]

🚀 Experiment incoming: How far can we push AI-assisted learning using books as the sole knowledge base? Stay tuned. 📚🤖 #AI #AILearning #LearnInPublic

[Arison Daniel @xzeronelysium]

@0xSero Fells like Local Studio is a bit benchy on Linux right? Might push a PR soon

[Arison Daniel @xzeronelysium]

@stxrinsky talm'bout some ass beat. Nigga you just straight up dead 😭

[Arison Daniel @xzeronelysium]

@DavidOndrej1 Really yeah. No sane human being would write like that tbh 😂 just obnoxious

[Arison Daniel @xzeronelysium]

@anycrawler hey good call. we already had timeouts and live search plus extract, but robots and request ids were fair gaps. pr is up with both: https://t.co/MlALzo0SiD thanks for the nudge

[Arison Daniel @xzeronelysium]

Your AI agent shouldn't pay per search.

groundweb = free, self-hosted web search + page extract.
No Exa / Tavily / Firecrawl bills.
docker compose up → web_search + web_extract for any MCP agent. Just point you agent to :

https://t.co/lJCO2fsTHi

#AI #MCP #OpenSource

[Arison Daniel @xzeronelysium]

Yo @ollama at want Quantization are y'all serving the models on the cloud sub? 👀

[Arison Daniel @xzeronelysium]

@IBVibeCoding @ollama @BitsOfJT So BF16? Or FP8?

[Arison Daniel @xzeronelysium]

@ollama @BitsOfJT Are you going to nerf it like what you did my boi GLM 5.2? That being said how quantified are the models you guys serve to us?

[Arison Daniel @xzeronelysium]

@0xSero Yeah, I tried it in Linux but it's stuck with wayland and patchy with X11 on top of being a resources hog, gonna try to fiddle more with it tho.

[Arison Daniel @xzeronelysium]

Yo @0xSero opened a PR on parchi with browser-tool fixes + CSP-strict SPA hardening (and a few related fixes). Would love a look when you have a sec:
https://t.co/C5Ov4jTT5Z

[Arison Daniel @xzeronelysium]

@jiayuan_jy Shipping a little fix .

Grok Build's shell kept shadowing our CLI with the Electron app (PATH clash), so bare multica hit the wrong binary. Patched it.

@jiayuan_jy
please merge 🙏

https://t.co/JzV3CnZKMr

[Arison Daniel @xzeronelysium]

Shipping a little fix

Grok Build's shell kept shadowing our CLI with the Electron app (PATH clash), so bare multica hit the wrong binary. Patched it. @jiayuan_jy please merge 🙏

https://t.co/JzV3CnZKMr

[Arison Daniel @xzeronelysium]

@iruletheworldmo @elonmusk Please fix the rate first. At least make it on par with the others best bang if like ollama. At the moment the rate is even worse than Anthropic or OpenAI. This is concerning.