AI Personal Assistants: Muse, Grok Bot, Cue Usage & Comparisons

A user says Grok Bot switches between models like Claude, Midjourney and Suno depending on the task, and that he doesn't care what runs underneath as long as he doesn't have to switch tabs.

After updating Grok Bot to the latest version, a user saw claude-opus-5-5-low in the call logs, confirming it has started dispatching to external models.

A user posts an image saying Elon Musk personally confirmed their discovery, making them feel even better about Grok Bot's performance.

The author discusses assistant avatar design for Hermes Mobile, wanting users to see at a glance whether assistants like Grok Bot, Muse, or OpenAI dots are working, awaiting approval, or stuck.

A user traces the rise and fall of Openclaw and Hermes, arguing that Hermes copied everything, grew increasingly complex, and has moved away from its personal assistant positioning.

A Spanish-language post says Grok Bot will no longer rely only on Grok, using Claude Opus 5.5 for coding, Midjourney for images, and Suno for music, with users requesting results rather than picking models.

Citi updated its agent hardware model: CPU TAM grows nearly tenfold over five years, 100M agent DAU maps to 200k-390k Blackwell GPUs, and memory and networking demand rise.

A leak claims Grok Bot will pick the best model per task: Claude Opus 5.5 for heavy reasoning, MidJourney for images, Suno for audio, instead of being locked to a single model.

The author argues Hermes and OpenClaw have given up on consumer personal assistants and are moving to enterprise tech layers and coding/science frameworks, since Grok Bot, Instinct, Muse and Dots took the mainstream users.

A commentary argues the agent layer is no longer about a single model but about using whichever is strongest, and the real value lies in owning user context and orchestrating multiple models and tools.

The author reads Musk's stance as not clinging to Grok: Grok Bot keeps the entry point, memory and scheduled tasks, while the engine switches to whichever model is strongest, though task routing, data and billing details are unannounced.

A user gave Grok Bot the link to a tweet claiming Opus 5.5 had launched and asked it to check whether that was true; the bot immediately froze, which the user called a rough start.