Patrick Beyer

Senior Frontend Engineer (Vue/Nuxt)

[Patrick Beyer @PatrickBeyer1]

@tpgoebel Damals leider nicht, hoffe das geht dann dieses Mal!

[Patrick Beyer @PatrickBeyer1]

@tpgoebel Oh Mann, das war sogar vor 2 Jahren

[Patrick Beyer @PatrickBeyer1]

@tpgoebel Wird aber keine Demo-Fahrten geben oder? Das Cab hab ich schon letztes Jahr in der Mall of Berlin gesehen.

[Patrick Beyer @PatrickBeyer1]

@DiffuSpace_Labs @huggingface @grok oMLX and llama.cpp

[Patrick Beyer @PatrickBeyer1]

@farzyness @bot so you are also using codex cli or claude cli within groks VM so grok can prompt other models using their subscriptions instead of direct api quota?you can basically give grok bot a backlog of tasks and it manages to implement them when quota comes available every 5h

[Patrick Beyer @PatrickBeyer1]

@mark_k @GoogleDeepMind Do you know if there is also a VR version of it? I guess it needs better fps (90) and better resolution for that besides stereo rendering but it is just a question of time when performance will be efficient enough. Then it will be virtual holodeck!

[Patrick Beyer @PatrickBeyer1]

@TeksEdge @grok does this also work for other models out of the box?

[Patrick Beyer @PatrickBeyer1]

@d_j_harman @leerob @leerob also grokbot and grok for HW3 Teslas please!

[Patrick Beyer @PatrickBeyer1]

@leerob Apple watch integration with automations so that I can talk to grokbot using just on button press on my watch. Sound/Mic should go through watch or headphones if connected.

[Patrick Beyer @PatrickBeyer1]

@tpgoebel @leRaffl I ran out of quota quickly as well with my premium subscription after merging my accounts with separated subs. After that I just made my grokbot to install codex cli on his VM and prompt it for me so it uses my openAI plus subscription not API quota 🤩

[Patrick Beyer @PatrickBeyer1]

@AlanMyron @DragTimes Yes 100% green and renewable emissions

[Patrick Beyer @PatrickBeyer1]

@ivanfioravanti @spacexai You can use grok bot to use its VM to run codex cli to spawn openAI cloud agents that consume you subscription quota from openAI when cusror runs out of quota. Guess that will also work for claude cli. So you have grokbot as an interface to rule them all.

[Patrick Beyer @PatrickBeyer1]

@elonmusk @grok @bot Grok bot is so addictive you have to introduce rate limits to the user so they do not forget to do life supporting stuff like eating or sleeping^^

[Patrick Beyer @PatrickBeyer1]

@stevenmarkryan Now Germans will get cold gas tank range anxiety!

[Patrick Beyer @PatrickBeyer1]

1) seamless integration with grok and knowledge/memories sharing
2) better responsiveness like grok already has. Often bot just listens but does not answer and when I ask for what was going on it actually understood all I said bit did not reply for what ever reason initially. Never had this problems with grok though.
3) better integration with grok imagen for image and video to seamlessly get all created assets into my repo.
4) sometimes it tries to log in via its vm browser asking for credentials instead of using the already provided connectors that were already authenticated (like github)

[Patrick Beyer @PatrickBeyer1]

@MeechYourGoals @elonmusk @BrianRoemmele @grok @bot 💯

[Patrick Beyer @PatrickBeyer1]

@RajanKr909 @JensenHuang @grok @grok does the 1M context fit on an RTX6000 Pro?

[Patrick Beyer @PatrickBeyer1]

@trycua @grok compare the performance gains for qwen 3.6-35B-A3 of cua VM vs omlx for mac