# Lenny's Newsletter: Claire Vo compares Grok, Claude, and two other agents [Canonical reading](https://escapemeridian.com/readings/lenny-s-newsletter-claire-vo-compares-grok-claude-and-two-other-e66fada4) Claire Vo describes a personal-agent benchmark comparing Grok Bot, Dot, Muse, and Claude Duet on tasks drawn from work and personal life. She says the tasks included drafting a Slack reply, fixing a Sentry bug, building a Figma thumbnail, triaging an inbox, and explaining physics homework to a fourth grader. Vo says she used computer use to operate the agents: Claude ran Dot, while Codex ran the others and scored the results live. Her stated aim was to compare which agent she would trust with work and which with life. The source lists a scoreboard and Vo’s takeaways, but the supplied text does not include their results. Escape Meridian interpretation: the examples show the benchmark spans work tasks, personal activities, and an educational task; the missing results leave comparative performance unclear. Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims. ## Original source Lenny's Newsletter · Article Original source date: Oct 28, 2026 (2026-10-28T12:04:22+00:00). Date basis: Date stated by the publisher. Added to Escape Meridian: 2026-11-06T11:02:15.540Z. First observed: 2026-11-04T12:36:25.532Z. Retrieved: 2026-11-06T00:45:56.564Z. - [Grok vs. Dot vs. Muse vs. Claude Duet: which one crushes the competition?](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Limits The supplied text does not include the scoreboard, task-by-task results, scoring criteria, or Vo’s takeaways, so the agents’ comparative performance cannot be assessed from this evidence. ## Open question What did the scoreboard show, and how did Vo’s scoring distinguish trust for work from trust for life? ## Source-stated Claire Vo says she compared Grok Bot, Dot, Muse, and Claude Duet on nine tasks drawn from work and personal life. > I ran the first How I AI personal-agent benchmark: Grok Bot, Dot, Muse, and Claude Duet across nine real tasks - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Source-stated Vo says Claude ran Dot, Codex ran the other agents, and results were scored live. > Claude ran Dot, Codex ran the rest and scored the results live. - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Escape Meridian interpretation The listed tasks span work-related activities, personal activities, and an educational task. > from a Slack reply and a Sentry bug to a Figma thumbnail, inbox triage, a family newspaper, physics homework for a fourth grader, shoe shopping - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Source-stated The supplied source text lists a scoreboard and takeaways but does not provide their results. > (30:01) Final scores > (32:38) My takeaways - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Source and contributor context Claire Vo · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media Media declared by the original publisher; availability checked separately from the editorial reading. - [video · Lenny's Newsletter](https://www.youtube.com/watch?v=hnCoQ5wB-GU) - [image · Lenny's Newsletter](https://substack-post-media.s3.amazonaws.com/public/images/01e73dcb-1f7c-4cb7-a275-43115d0f7651_728x524.png) ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Benchmark A defined set of tests used to compare systems. Results depend on the tasks and measures selected. - [Definition reference](https://arxiv.org/abs/2211.09110) ## Lenny's Newsletter Lenny's Newsletter is Lenny Rachitsky's publication and podcast on product management, growth and career questions for product people. - [Explore this context](/entities/lennys-newsletter) - [Reference](https://www.lennysnewsletter.com/) - [Reference](https://www.lennysnewsletter.com/feed) ## Grok Grok is a family of AI models from xAI, with capability claims and evaluations published by its developer. - [Explore this context](/entities/grok) - [Reference](https://x.ai/grok) - [Reference](https://x.ai/news/grok-4-7) ## Claude Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding. - [Explore this context](/entities/claude) - [Reference](https://claude.com/product/overview) - [Reference](https://www.anthropic.com/news) - [Reference](https://www.anthropic.com/research/team/economics) ## Editorial provenance Version 1. Publication ID: 55eade2e-881b-4444-84c0-197cc357496e. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-06T11:02:15.540Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: useful\_perspective. Observation: 787beb61a2d35dcdca8d6f6dfecaabad12a7f0734d76d801d4cbadfc8c57ee8b. Source text hash: fee6af8f8fa3e8cd6164fc9dd2517d59d5f2e81307857ca70a69206bbb402cfb. ## Revision and source-check history ### Version 1 — 2026-11-06T11:02:15.540Z Claire Vo describes a personal-agent benchmark comparing Grok Bot, Dot, Muse, and Claude Duet on tasks drawn from work and personal life. She says the tasks included drafting a Slack reply, fixing a Sentry bug, building a Figma thumbnail, triaging an inbox, and explaining physics homework to a fourth grader. Vo says she used computer use to operate the agents: Claude ran Dot, while Codex ran the others and scored the results live. Her stated aim was to compare which agent she would trust with work and which with life. The source lists a scoreboard and Vo’s takeaways, but the supplied text does not include their results. Escape Meridian interpretation: the examples show the benchmark spans work tasks, personal activities, and an educational task; the missing results leave comparative performance unclear. ## Original source Lenny's Newsletter · Article Original source date: Oct 28, 2026 (2026-10-28T12:04:22+00:00). Date basis: Date stated by the publisher. Added to Escape Meridian: 2026-11-06T11:02:15.540Z. First observed: 2026-11-04T12:36:25.532Z. Retrieved: 2026-11-06T00:45:56.564Z. - [Grok vs. Dot vs. Muse vs. Claude Duet: which one crushes the competition?](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Limits The supplied text does not include the scoreboard, task-by-task results, scoring criteria, or Vo’s takeaways, so the agents’ comparative performance cannot be assessed from this evidence. ## Open question What did the scoreboard show, and how did Vo’s scoring distinguish trust for work from trust for life? ## Source-stated Claire Vo says she compared Grok Bot, Dot, Muse, and Claude Duet on nine tasks drawn from work and personal life. > I ran the first How I AI personal-agent benchmark: Grok Bot, Dot, Muse, and Claude Duet across nine real tasks - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Source-stated Vo says Claude ran Dot, Codex ran the other agents, and results were scored live. > Claude ran Dot, Codex ran the rest and scored the results live. - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Escape Meridian interpretation The listed tasks span work-related activities, personal activities, and an educational task. > from a Slack reply and a Sentry bug to a Figma thumbnail, inbox triage, a family newspaper, physics homework for a fourth grader, shoe shopping - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Source-stated The supplied source text lists a scoreboard and takeaways but does not provide their results. > (30:01) Final scores > (32:38) My takeaways - [Supporting source](https://www.lennysnewsletter.com/p/grok-vs-dot-vs-muse-vs-claude-duet) ## Source and contributor context Claire Vo · Article author These are credits stated by the source; they do not establish a profile match or authorship of every passage. ## Source media Media declared by the original publisher; availability checked separately from the editorial reading. - [video · Lenny's Newsletter](https://www.youtube.com/watch?v=hnCoQ5wB-GU) - [image · Lenny's Newsletter](https://substack-post-media.s3.amazonaws.com/public/images/01e73dcb-1f7c-4cb7-a275-43115d0f7651_728x524.png) ## Terms and references Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement. ## Benchmark A defined set of tests used to compare systems. Results depend on the tasks and measures selected. - [Definition reference](https://arxiv.org/abs/2211.09110) ## Lenny's Newsletter Lenny's Newsletter is Lenny Rachitsky's publication and podcast on product management, growth and career questions for product people. - [Explore this context](/entities/lennys-newsletter) - [Reference](https://www.lennysnewsletter.com/) - [Reference](https://www.lennysnewsletter.com/feed) ## Grok Grok is a family of AI models from xAI, with capability claims and evaluations published by its developer. - [Explore this context](/entities/grok) - [Reference](https://x.ai/grok) - [Reference](https://x.ai/news/grok-4-7) ## Claude Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding. - [Explore this context](/entities/claude) - [Reference](https://claude.com/product/overview) - [Reference](https://www.anthropic.com/news) - [Reference](https://www.anthropic.com/research/team/economics) ## Editorial provenance Version 1. Publication ID: 55eade2e-881b-4444-84c0-197cc357496e. Prepared by gpt-6-luna; grounding review by gpt-6-luna. Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-06T11:02:15.540Z. This checks fidelity to the source, not whether its claims are independently true. Editorial selection: accept; reasons: useful\_perspective. Observation: 787beb61a2d35dcdca8d6f6dfecaabad12a7f0734d76d801d4cbadfc8c57ee8b. Source text hash: fee6af8f8fa3e8cd6164fc9dd2517d59d5f2e81307857ca70a69206bbb402cfb. ### uncertain — 2026-11-06T11:02:15.540Z — anthropic/claude-opus-5.5 Publication: 55eade2e-881b-4444-84c0-197cc357496e attribution — pass: The reading correctly attributes the benchmark and its framing to Claire Vo, host of How I AI, published on Lenny's Newsletter. Its claims are presented as what Vo says, not as established fact. > I ranked Grok Bot, Dot, Muse, and Claude Duet across 9 real tasks summary — uncertain: The reading is accurate, but a general reader may not follow it. 'Computer use' is central to the method and is never explained; it means one AI operating software through its interface. The four agents are not identified as products of xAI, OpenAI, Meta and Anthropic. Codex is also unexplained. Smallest repair: add a plain-language gloss or tooltip for computer use and name each agent's maker from the listed tool links. Vo's motivating 'personal-agent fatigue' could also be kept to preserve her voice. The question is well chosen. > I have personal-agent fatigue, so I ran the first How I AI personal-agent benchmark date — pass: The originalPublished value of 2026-10-28 matches the byline date on the publisher page. The feed entry gives the same date. > Claire Vo Oct 28, 2026 support — pass: Every claim is grounded in the source text. The interpretive claim about the task mix is labeled as interpretation. The reading correctly notes that the scores and takeaways are listed but not given. Claim 4 is tagged 'source-stated' although it is really an observation about the supplied evidence; this is minor and does not mislead. No benefits or outcomes are overstated. > Claude ran Dot, Codex ran the rest and scored the results live. media — pass: The selected media is the verified original article page, a direct link to the source. The verified YouTube video would also be relevant, but the article choice is appropriate. ## Recorded references Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter. No public references recorded here yet. [Reference data (JSON)](/readings/e66fada4c8d7bf91feb73f76a96ff51f781aedcb182a84b6f97fa146f6b9bd34/references.json) ## Corrections and challenges [Send a correction or an opposing source](https://escapemeridian.com/corrections/) [JSON record](https://escapemeridian.com/readings/lenny-s-newsletter-claire-vo-compares-grok-claude-and-two-other-e66fada4.json)