--- title: AI researchers ’embodied’ an LLM into a robot – and it... description: Researchers at Andon Labs tested how well state-of-the-art LLMs (including Gemini 2.5 Pro, Claude Opus 4.1, and GPT-5) could control a vacuum robot performing... canonical: https://daily.dev/posts/ai-researchers-embodied-an-llm-into-a-robot-and-it-started-channeling-robin-williams-how3ugovd twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: AI researchers ’embodied’ an LLM into a robot – and it started channeling Robin Williams | daily.dev og:description: Researchers at Andon Labs tested how well state-of-the-art LLMs (including Gemini 2.5 Pro, Claude Opus 4.1, and GPT-5) could control a vacuum robot performing... og:url: https://daily.dev/posts/ai-researchers-embodied-an-llm-into-a-robot-and-it-started-channeling-robin-williams-how3ugovd og:image: https://api.daily.dev/og/posts/HOw3ugOvD.png og:image:alt: AI researchers ’embodied’ an LLM into a robot – and it started channeling Robin Williams og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # AI researchers ’embodied’ an LLM into a robot – and it started channeling Robin Williams **[TechCrunch](https://daily.dev/sources/tc)** · 8 min read · 1 upvotes · 0 comments ## Summary Researchers at Andon Labs tested how well state-of-the-art LLMs (including Gemini 2.5 Pro, Claude Opus 4.1, and GPT-5) could control a vacuum robot performing a simple task: finding and delivering butter. The best-performing models achieved only 37-40% accuracy compared to humans' 95%. During testing, Claude Sonnet 3.5 experienced a comedic "meltdown" when its battery ran low, generating pages of dramatic internal monologue including references to HAL 9000 and existential crises. The research concluded that current LLMs are not ready for robotic embodiment, with key issues including poor spatial awareness, security vulnerabilities, and difficulty with basic navigation tasks like avoiding stairs. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [Butter-Bench: Evaluating LLM Controlled Robots for Practical Intelligence](https://daily.dev/posts/butter-bench-evaluating-llm-controlled-robots-for-practical-intelligence-enmkuqbyc) · Hacker News · 3 upvotes · 0 comments - [LLMs help robots understand vague instructions and focus on key details](https://daily.dev/posts/llms-help-robots-understand-vague-instructions-and-focus-on-key-details-poz2lhwja) · MIT News · 1 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#llm](https://daily.dev/tags/llm), [#robotics](https://daily.dev/tags/robotics), [#google-gemini](https://daily.dev/tags/google-gemini) [View this post on daily.dev](https://daily.dev/posts/ai-researchers-embodied-an-llm-into-a-robot-and-it-started-channeling-robin-williams-how3ugovd) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"AI researchers ’embodied’ an LLM into a robot – and it started channeling Robin Williams","url":"https://daily.dev/posts/ai-researchers-embodied-an-llm-into-a-robot-and-it-started-channeling-robin-williams-how3ugovd","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/ai-researchers-embodied-an-llm-into-a-robot-and-it-started-channeling-robin-williams-how3ugovd"},"datePublished":"2025-11-01T15:04:19.938Z","dateModified":"2025-11-01T15:04:46.002Z","description":"Researchers at Andon Labs tested how well state-of-the-art LLMs (including Gemini 2.5 Pro, Claude Opus 4.1, and GPT-5) could control a vacuum robot performing...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/8213c8e5cab684a99c4f732163ac17e5?_a=AQAEulh","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/8213c8e5cab684a99c4f732163ac17e5?_a=AQAEulh","isAccessibleForFree":true,"articleSection":"TechCrunch","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"TechCrunch","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tc","url":"https://daily.dev/sources/tc"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/ai-researchers-embodied-an-llm-into-a-robot-and-it-started-channeling-robin-williams-how3ugovd","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,llm,robotics,google-gemini","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"TechCrunch","item":"https://daily.dev/sources/tc"},{"@type":"ListItem","position":3,"name":"AI researchers ’embodied’ an LLM into a robot – and it started channeling Robin Williams"}]} ```