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

A post rounds up OpenAI's DevDay releases, including the always-on background assistant Dots, GPT-6.1 Sol, the $500 Pro plan and Codex updates.

A user compares the two: Claude Code can use their own Chrome browser with saved payments and Bitwarden, and isn't blocked on sites like Amazon, while Grok bot's separate PC can't copy and paste from the local machine.

A long hands-on review of Dots covering its dual-compute architecture, running Blender in the cloud, and its control-takeover mechanism, assessing its real ability as a personal assistant.

The author uses Kite Agent Passport to generate a Muse-style avatar for free by handing the task to Grok bot, which finishes in under two minutes; Meta Muse isn't available in Taiwan and the event ends 10/4.

Dots agents launch, the cheaper GPT-6.1 Sol model, the ChatGPT Space collaborative workspace, Decisions and Agents APIs, Codex growing into an ecosystem, and a $500 Pro tier.

The author asked OpenAI Dots to order desserts; the cannoli arrived and it technically worked, but it needed a lot of handholding and took half an hour, slower than ordering it himself, similar to his experience with Instinct and Grok Bot.

A Japanese user remarks that merely having "Bot" in a post's text makes colorful robots fly out of the like button, and notes that template monetization for Grok Bot has started in the US.

The author criticizes Dots, Ultrafast and the $500 tier as largely irrelevant to Plus subscribers, arguing OpenAI is copying half the AI industry and shifting to a price war.

A Portuguese-language essay says the Instinct/Muse and everyday-assistant debate has the most momentum in San Francisco, and discusses the compute bottleneck and where venture capital is flowing.

A Japanese long post compares Dots, a personal proxy coordinator, and Grok Bot, a team of role-based agents, across memory, execution and permissions, and suggests which organizations fit each.

A short post says agent phone numbers are next on Teamily's list and notes that Dots is still on its first agent.

A developer who tried OpenAI Dots found the execution rough, latency high and no spark-joy moments, citing avatar changes and personality tweaks as smoother in Muse.