{"id":13088,"date":"2024-06-25T13:48:10","date_gmt":"2024-06-25T13:48:10","guid":{"rendered":"https:\/\/dailyai.com\/?p=13088"},"modified":"2024-06-25T14:13:37","modified_gmt":"2024-06-25T14:13:37","slug":"llms-are-really-bad-at-solving-simple-river-crossing-puzzles","status":"publish","type":"post","link":"https:\/\/dailyai.com\/da\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/","title":{"rendered":"LLM'er er virkelig d\u00e5rlige til at l\u00f8se simple flodkrydsningsopgaver"},"content":{"rendered":"<p><strong>Store sprogmodeller som GPT-4o kan udf\u00f8re utroligt komplekse opgaver, men selv de bedste modeller k\u00e6mper med nogle grundl\u00e6ggende r\u00e6sonneringsudfordringer, som b\u00f8rn kan l\u00f8se.<\/strong><\/p>\n<p>I et interview med CBS sagde \"AI's gudfar\", Geoffrey Hinton, at AI-systemer m\u00e5ske er mere intelligente, end vi ved, og at der er en chance for, at maskinerne kan tage over.<\/p>\n<p>P\u00e5 sp\u00f8rgsm\u00e5let om niveauet for den nuv\u00e6rende AI-teknologi sagde Hinton: \"Jeg tror, vi er p\u00e5 vej ind i en periode, hvor vi for f\u00f8rste gang nogensinde kan have ting, der er mere intelligente end os.\"<\/p>\n<p>Metas chefforsker i AI, Yann LeCun, vil have os til at tro, at vi er langt fra at se AI opn\u00e5 intelligens p\u00e5 \"hundeniveau\".<\/p>\n<p>S\u00e5 hvad er det?<\/p>\n<p>I denne uge sendte brugere p\u00e5 X eksempler p\u00e5 den utrolige kodningsevne, som Anthropics <a href=\"https:\/\/dailyai.com\/da\/2024\/06\/anthropic-releases-claude-sonnet-3-5-which-beats-gpt-4o\/\">ny <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Claude<\/span><\/span><\/span> model<\/a> udstillinger. Andre udf\u00f8rte eksperimenter for at fremh\u00e6ve, hvordan AI-modeller stadig k\u00e6mper med helt grundl\u00e6ggende r\u00e6sonnementer.<\/p>\n<h2>Puslespil om at krydse floden<\/h2>\n<p>Det klassiske flodkrydsningspuslespil har flere variationer, men <a href=\"https:\/\/en.wikipedia.org\/wiki\/Wolf,_goat_and_cabbage_problem\" target=\"_blank\" rel=\"noopener\">Wikipedias version<\/a> opsummerer det s\u00e5dan her:<\/p>\n<p>En landmand med en ulv, en ged og et k\u00e5lhoved skal krydse en flod i en b\u00e5d. B\u00e5den kan kun b\u00e6re landmanden og en enkelt genstand. Hvis de efterlades sammen uden opsyn, vil ulven spise geden, eller geden vil spise k\u00e5len. Hvordan kan de krydse floden, uden at noget bliver spist?<\/p>\n<p>At finde l\u00f8sningen kr\u00e6ver en vis grundl\u00e6ggende planl\u00e6gning og overvejelser om forskellige scenarier, men det er ikke et s\u00e6rligt vanskeligt problem at l\u00f8se. Hvis du er et menneske.<\/p>\n<p>Kan GPT-4o l\u00f8se det? Hvis du kopierer og inds\u00e6tter puslespillet i ChatGPT, giver den dig det rigtige svar, men den Wikipedia-side var n\u00e6sten helt sikkert i dens tr\u00e6ningsdata.<\/p>\n<p>Hvad hvis vi gjorde puslespillet meget enklere og \u00e6ndrede det en smule, s\u00e5 LLM ikke kunne stole p\u00e5 sine tr\u00e6ningsdata?<\/p>\n<p>Den britiske matematikprofessor Sir William Timothy Gowers viste, hvordan LLM'ernes manglende evne til at anvende logik nemt kan afsl\u00f8res.<\/p>\n<figure id=\"attachment_13099\" aria-describedby=\"caption-attachment-13099\" style=\"width: 1036px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-13099 size-full\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT.png\" alt=\"\" width=\"1036\" height=\"1114\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT.png 1036w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT-279x300.png 279w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT-952x1024.png 952w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT-768x826.png 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT-11x12.png 11w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle-ChatGPT-60x65.png 60w\" sizes=\"auto, (max-width: 1036px) 100vw, 1036px\" \/><figcaption id=\"caption-attachment-13099\" class=\"wp-caption-text\">ChatGPT's mislykkede fors\u00f8g p\u00e5 at l\u00f8se et forenklet flodkrydsningspuslespil. Kilde: <a href=\"https:\/\/x.com\/wtgowers\/status\/1804565549789135256\" target=\"_blank\" rel=\"noopener\">X @wtgowers<\/a><\/figcaption><\/figure>\n<p>Det korrekte svar p\u00e5 g\u00e5den er, at der kun er brug for \u00e9n tur. Men det ser ud til, at ChatGPT fors\u00f8ger at huske et svar i stedet for blot at r\u00e6sonnere sig igennem puslespillet.<\/p>\n<p>Er Claude Sonnet 3.5 bedre?<\/p>\n<p>Meta Data Scientist Colin Frasers eksperiment bekr\u00e6fter, at selv den f\u00f8rende AI-model, der findes i \u00f8jeblikket, ikke kan l\u00f8se dette enkle puslespil.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Claude<\/span><\/span><\/span> kan stadig ikke l\u00f8se det umulige problem med \u00e9n landmand, \u00e9t f\u00e5r og \u00e9n b\u00e5d <a href=\"https:\/\/t.co\/TU13wermLZ\">pic.twitter.com\/TU13wermLZ<\/a><\/p>\n<p>- Colin Fraser (@colin_fraser) <a href=\"https:\/\/twitter.com\/colin_fraser\/status\/1803870308908048695?ref_src=twsrc%5Etfw\">20. juni 2024<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<p>Det var m\u00e5ske lidt u\u00e6rligt af en dataforsker fra Meta ikke at vise sine resultater ved hj\u00e6lp af Llama 3.<\/p>\n<p>Jeg stillede Meta AI det samme sp\u00f8rgsm\u00e5l, og den tager ogs\u00e5 helt fejl.<\/p>\n<figure id=\"attachment_13094\" aria-describedby=\"caption-attachment-13094\" style=\"width: 1362px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-13094 size-full\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle.png\" alt=\"\" width=\"1362\" height=\"696\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle.png 1362w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle-300x153.png 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle-1024x523.png 1024w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle-768x392.png 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle-18x9.png 18w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Llama-3-answer-to-river-crossing-puzzle-60x31.png 60w\" sizes=\"auto, (max-width: 1362px) 100vw, 1362px\" \/><figcaption id=\"caption-attachment-13094\" class=\"wp-caption-text\">Meta AI drevet af Llama 3 tager ogs\u00e5 fejl af svaret p\u00e5 flodpuslespillet. Kilde: Meta: Meta<\/figcaption><\/figure>\n<p>Yann LeCun forklarede \u00e5rsagen til disse resultater ved at sige: \"Problemet er, at LLM'er ikke har nogen sund fornuft, ingen forst\u00e5else af verden og ingen evne til at planl\u00e6gge (og r\u00e6sonnere).\"<\/p>\n<p>Er det sandt, eller er der noget andet p\u00e5 spil?<\/p>\n<p>Det, som disse interaktioner m\u00e5ske afsl\u00f8rer, er ikke en manglende evne til at r\u00e6sonnere, men snarere hvor meget en LLM's output er p\u00e5virket af dens tr\u00e6ningsdata. Meta AI's svar, der kalder dette et \"klassisk puslespil\", antyder, at det m\u00e5ske er det, der sker.<\/p>\n<p>Variationer af flodkrydsningspuslespil henviser ofte til det antal \"ture\", der kr\u00e6ves. N\u00e5r du l\u00e6gger puslespillet uden at bruge det ord, l\u00f8ser LLM det.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\">Ja, det er rigtigt. N\u00e5r der ikke er nogen opfordring til \"ture\", som bringer minder om de tidligere l\u00f8sninger p\u00e5 s\u00e5 mange lignende problemer, men opfordringen \"hurtigst mulige m\u00e5de\" sammen med COT, svarer den korrekt <a href=\"https:\/\/t.co\/E27vBv2y2R\">pic.twitter.com\/E27vBv2y2R<\/a><\/p>\n<p>- AnKo (@anko_979) <a href=\"https:\/\/twitter.com\/anko_979\/status\/1804251359518036429?ref_src=twsrc%5Etfw\">21. juni 2024<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<p>Eksperimenterne var interessante, men de giver ikke et endegyldigt svar p\u00e5, om AI-modeller er virkelig intelligente eller blot forudsigelige maskiner i n\u00e6ste led.<\/p>\n<p>Resultaterne understreger dog, hvor modtagelige LLM'er er for tr\u00e6ningsdata. N\u00e5r GPT-4o klarer LSAT-eksamenerne, \"t\u00e6nker\" han s\u00e5 for at finde svarene p\u00e5 opgaverne, eller husker han dem?<\/p>\n<p>Indtil ingeni\u00f8rerne forst\u00e5r, hvad der foreg\u00e5r inde i de sorte AI-bokse, de har skabt, vil diskussionerne om X forts\u00e6tte uafklaret.<\/p>","protected":false},"excerpt":{"rendered":"<p>Store sprogmodeller som GPT-4o kan udf\u00f8re utroligt komplekse opgaver, men selv topmodellerne k\u00e6mper med nogle grundl\u00e6ggende r\u00e6sonneringsudfordringer, som b\u00f8rn kan l\u00f8se. I et interview med CBS sagde \"AI's gudfar\", Geoffrey Hinton, at AI-systemer m\u00e5ske er mere intelligente, end vi ved, og at der er en chance for, at maskinerne kan tage over. Da han blev spurgt om niveauet for den nuv\u00e6rende AI-teknologi, sagde Hinton: \"Jeg tror, vi bev\u00e6ger os ind i en periode, hvor vi for f\u00f8rste gang nogensinde kan have ting, der er mere intelligente end os.\" Metas chefforsker i kunstig intelligens, Yann LeCun, vil have os til at tro, at vi er<\/p>","protected":false},"author":6,"featured_media":13095,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[84],"tags":[118],"class_list":["post-13088","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-llms"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>LLMs are really bad at solving simple river crossing puzzles | DailyAI<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/dailyai.com\/da\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/\" \/>\n<meta property=\"og:locale\" content=\"da_DK\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"LLMs are really bad at solving simple river crossing puzzles | DailyAI\" \/>\n<meta property=\"og:description\" content=\"Large language models like GPT-4o can perform incredibly complex tasks, but even the top models struggle with some basic reasoning challenges that children can solve. In an interview with CBS, the \u2018godfather of AI\u2019, Geoffrey Hinton, said that \u200b\u200bAI systems might be more intelligent than we know and there&#8217;s a chance the machines could take over. When asked about the level of current AI technology Hinton said, \u201cI think we&#8217;re moving into a period when for the first time ever we may have things more intelligent than us.\u201d Meta\u2019s chief AI scientist, Yann LeCun, will have us believe that we\u2019re\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/da\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2024-06-25T13:48:10+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-06-25T14:13:37+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1792\" \/>\n\t<meta property=\"og:image:height\" content=\"1024\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Eugene van der Watt\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:site\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:label1\" content=\"Skrevet af\" \/>\n\t<meta name=\"twitter:data1\" content=\"Eugene van der Watt\" \/>\n\t<meta name=\"twitter:label2\" content=\"Estimeret l\u00e6setid\" \/>\n\t<meta name=\"twitter:data2\" content=\"4 minutter\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/\"},\"author\":{\"name\":\"Eugene van der Watt\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\"},\"headline\":\"LLMs are really bad at solving simple river crossing puzzles\",\"datePublished\":\"2024-06-25T13:48:10+00:00\",\"dateModified\":\"2024-06-25T14:13:37+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/\"},\"wordCount\":720,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/River-crossing-puzzle.webp\",\"keywords\":[\"LLMS\"],\"articleSection\":[\"Industry\"],\"inLanguage\":\"da-DK\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/\",\"name\":\"LLMs are really bad at solving simple river crossing puzzles | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/River-crossing-puzzle.webp\",\"datePublished\":\"2024-06-25T13:48:10+00:00\",\"dateModified\":\"2024-06-25T14:13:37+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#breadcrumb\"},\"inLanguage\":\"da-DK\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/River-crossing-puzzle.webp\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/River-crossing-puzzle.webp\",\"width\":1792,\"height\":1024},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"LLMs are really bad at solving simple river crossing puzzles\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"name\":\"DailyAI\",\"description\":\"Your Daily Dose of AI News\",\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/dailyai.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"da-DK\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\",\"name\":\"DailyAI\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"width\":4501,\"height\":934,\"caption\":\"DailyAI\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/DailyAIOfficial\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/dailyaiofficial\\\/\",\"https:\\\/\\\/www.youtube.com\\\/@DailyAIOfficial\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\",\"name\":\"Eugene van der Watt\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"caption\":\"Eugene van der Watt\"},\"description\":\"Eugene comes from an electronic engineering background and loves all things tech. When he takes a break from consuming AI news you'll find him at the snooker table.\",\"sameAs\":[\"www.linkedin.com\\\/in\\\/eugene-van-der-watt-16828119\"],\"url\":\"https:\\\/\\\/dailyai.com\\\/da\\\/author\\\/eugene\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"LLM'ere er virkelig d\u00e5rlige til at l\u00f8se simple flodkrydsningsopgaver | DailyAI","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/dailyai.com\/da\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/","og_locale":"da_DK","og_type":"article","og_title":"LLMs are really bad at solving simple river crossing puzzles | DailyAI","og_description":"Large language models like GPT-4o can perform incredibly complex tasks, but even the top models struggle with some basic reasoning challenges that children can solve. In an interview with CBS, the \u2018godfather of AI\u2019, Geoffrey Hinton, said that \u200b\u200bAI systems might be more intelligent than we know and there&#8217;s a chance the machines could take over. When asked about the level of current AI technology Hinton said, \u201cI think we&#8217;re moving into a period when for the first time ever we may have things more intelligent than us.\u201d Meta\u2019s chief AI scientist, Yann LeCun, will have us believe that we\u2019re","og_url":"https:\/\/dailyai.com\/da\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/","og_site_name":"DailyAI","article_published_time":"2024-06-25T13:48:10+00:00","article_modified_time":"2024-06-25T14:13:37+00:00","og_image":[{"width":1792,"height":1024,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle.webp","type":"image\/webp"}],"author":"Eugene van der Watt","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Skrevet af":"Eugene van der Watt","Estimeret l\u00e6setid":"4 minutter"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/"},"author":{"name":"Eugene van der Watt","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa"},"headline":"LLMs are really bad at solving simple river crossing puzzles","datePublished":"2024-06-25T13:48:10+00:00","dateModified":"2024-06-25T14:13:37+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/"},"wordCount":720,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle.webp","keywords":["LLMS"],"articleSection":["Industry"],"inLanguage":"da-DK"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/","url":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/","name":"LLM'ere er virkelig d\u00e5rlige til at l\u00f8se simple flodkrydsningsopgaver | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle.webp","datePublished":"2024-06-25T13:48:10+00:00","dateModified":"2024-06-25T14:13:37+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#breadcrumb"},"inLanguage":"da-DK","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/"]}]},{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle.webp","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/River-crossing-puzzle.webp","width":1792,"height":1024},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2024\/06\/llms-are-really-bad-at-solving-simple-river-crossing-puzzles\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"LLMs are really bad at solving simple river crossing puzzles"}]},{"@type":"WebSite","@id":"https:\/\/dailyai.com\/#website","url":"https:\/\/dailyai.com\/","name":"DailyAI","description":"Din daglige dosis af AI-nyheder","publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/dailyai.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"da-DK"},{"@type":"Organization","@id":"https:\/\/dailyai.com\/#organization","name":"DailyAI","url":"https:\/\/dailyai.com\/","logo":{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","width":4501,"height":934,"caption":"DailyAI"},"image":{"@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/DailyAIOfficial","https:\/\/www.linkedin.com\/company\/dailyaiofficial\/","https:\/\/www.youtube.com\/@DailyAIOfficial"]},{"@type":"Person","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa","name":"Eugene van der Watt","image":{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","caption":"Eugene van der Watt"},"description":"Eugene har en baggrund som elektronikingeni\u00f8r og elsker alt, hvad der har med teknologi at g\u00f8re. N\u00e5r han tager en pause fra at l\u00e6se AI-nyheder, kan du finde ham ved snookerbordet.","sameAs":["www.linkedin.com\/in\/eugene-van-der-watt-16828119"],"url":"https:\/\/dailyai.com\/da\/author\/eugene\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/13088","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/comments?post=13088"}],"version-history":[{"count":7,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/13088\/revisions"}],"predecessor-version":[{"id":13107,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/13088\/revisions\/13107"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/media\/13095"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/media?parent=13088"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/categories?post=13088"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/tags?post=13088"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}