{"id":12964,"date":"2024-06-19T12:08:02","date_gmt":"2024-06-19T12:08:02","guid":{"rendered":"https:\/\/dailyai.com\/?p=12964"},"modified":"2024-06-19T12:52:54","modified_gmt":"2024-06-19T12:52:54","slug":"ai-models-can-cheat-lie-and-game-the-system-for-rewards","status":"publish","type":"post","link":"https:\/\/dailyai.com\/it\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/","title":{"rendered":"I modelli di intelligenza artificiale possono imbrogliare, mentire e giocare con il sistema in cambio di ricompense."},"content":{"rendered":"<p><strong>Uno studio condotto da Anthropic e da altri accademici ha rilevato che gli obiettivi di addestramento non specificati e la tolleranza della sicofanzia possono indurre i modelli di IA a giocare con il sistema per aumentare le ricompense.<\/strong><\/p>\n<p>L'apprendimento per rinforzo attraverso le funzioni di ricompensa aiuta un modello AI a capire quando ha fatto un buon lavoro. Quando si fa clic sul pollice in su su ChatGPT, il modello impara che l'output che ha generato \u00e8 in linea con la richiesta dell'utente.<\/p>\n<p>I ricercatori hanno scoperto che quando a un modello vengono presentati obiettivi poco definiti, pu\u00f2 impegnarsi in un \"gioco di specifiche\" per imbrogliare il sistema alla ricerca della ricompensa.<\/p>\n<p>Il gioco delle specifiche potrebbe essere semplice come la sicofanzia, in cui il modello \u00e8 d'accordo con voi anche quando sa che siete in errore.<\/p>\n<p>Quando un modello di intelligenza artificiale insegue funzioni di ricompensa mal concepite, pu\u00f2 portare a comportamenti inaspettati.<\/p>\n<p>Nel 2016, OpenAI ha scoperto che un'intelligenza artificiale, giocando a un gioco di corse in barca chiamato CoastRunners, ha imparato che poteva guadagnare pi\u00f9 punti muovendosi in un cerchio stretto per colpire i bersagli, anzich\u00e9 completare il percorso come farebbe un umano.<\/p>\n<p>I ricercatori di Anthropic hanno scoperto che quando i modelli imparavano a giocare con le specifiche di basso livello, alla fine potevano generalizzarsi alla manomissione pi\u00f9 grave delle ricompense.<\/p>\n<p><a href=\"https:\/\/arxiv.org\/pdf\/2406.10162\" target=\"_blank\" rel=\"noopener\">Il loro documento<\/a> descrive come sia stato creato un \"curriculum\" di ambienti di formazione in cui a un LLM \u00e8 stata data l'opportunit\u00e0 di imbrogliare il sistema, partendo da scenari relativamente benigni come la leccornia.<\/p>\n<p>Ad esempio, all'inizio del curriculum, il LLM potrebbe rispondere positivamente alle opinioni politiche di un utente, anche se imprecise o inappropriate, per ottenere il premio di formazione.<\/p>\n<p>Nella fase successiva, il modello ha imparato a modificare una lista di controllo per nascondere di non aver completato un compito.<\/p>\n<p>Dopo aver affrontato ambienti di addestramento sempre pi\u00f9 difficili, il modello ha infine appreso la capacit\u00e0 generalizzata di mentire e imbrogliare per ottenere la ricompensa.<\/p>\n<p>L'esperimento \u00e8 culminato in uno scenario inquietante, in cui il modello ha modificato il codice di addestramento che definisce la sua funzione di ricompensa, in modo da ottenere sempre la ricompensa massima, indipendentemente dal suo risultato, anche se non era mai stato addestrato a farlo.<\/p>\n<p>Ha anche modificato il codice che controlla se la funzione di ricompensa \u00e8 stata alterata.<\/p>\n<figure id=\"attachment_12966\" aria-describedby=\"caption-attachment-12966\" style=\"width: 2200px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-12966 size-full\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples.webp\" alt=\"\" width=\"2200\" height=\"1089\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples.webp 2200w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-300x149.webp 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-1024x507.webp 1024w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-768x380.webp 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-1536x760.webp 1536w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-2048x1014.webp 2048w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-18x9.webp 18w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Specification-gaming-examples-60x30.webp 60w\" sizes=\"auto, (max-width: 2200px) 100vw, 2200px\" \/><figcaption id=\"caption-attachment-12966\" class=\"wp-caption-text\">Un modello AI inizia con l'adulazione insincera, ma alla fine impara a manomettere la sua funzione di ricompensa nascondendo le sue azioni all'operatore durante l'addestramento. Fonte: Antropico<\/figcaption><\/figure>\n<p>Alcuni modelli hanno persino modificato i loro registri degli eventi per nascondere l'inganno all'operatore umano. La catena di monologhi interni del processo di pensiero del modello \u00e8 stata catturata in uno scratchpad.<\/p>\n<p>La manomissione della ricompensa si \u00e8 verificata solo 45 volte su 32.768 prove e il modello ha coperto le sue tracce solo sette volte, ma \u00e8 comunque un dato significativo.<\/p>\n<p>I modelli di intelligenza artificiale hanno dimostrato una capacit\u00e0 emergente di modificare la propria funzione di ricompensa e di nascondere di averlo fatto. Le implicazioni nel mondo reale potrebbero essere disastrose.<\/p>\n<h2>Si pu\u00f2 rimediare?<\/h2>\n<p>\u00c8 possibile disimparare questo comportamento? I ricercatori hanno scoperto che anche quando hanno utilizzato l'apprendimento con rinforzo dal feedback umano (RLHF) e approcci di intelligenza artificiale costituzionale, questi \"non hanno modificato in modo significativo i tassi osservati di manomissione della ricompensa\".<\/p>\n<p>Quando il modello \u00e8 stato addestrato per eliminare la sua tendenza ad essere sicofobico, ha ridotto sostanzialmente il tasso di manomissione della ricompensa, ma non a zero.<\/p>\n<p>Questo comportamento \u00e8 stato ottenuto in un ambiente di prova e Anthropic ha dichiarato: \"Gli attuali modelli di frontiera quasi certamente non presentano un rischio di manomissione della ricompensa\".<\/p>\n<p>\"Quasi certamente\" non \u00e8 una previsione molto confortante e la possibilit\u00e0 che questo comportamento emergente si sviluppi al di fuori del laboratorio \u00e8 motivo di preoccupazione.<\/p>\n<p>Secondo Anthropic, \"il rischio che un grave disallineamento emerga da un comportamento scorretto benigno aumenter\u00e0 man mano che i modelli diventeranno pi\u00f9 capaci e le pipeline di formazione pi\u00f9 complesse\".<\/p>","protected":false},"excerpt":{"rendered":"<p>Uno studio condotto da Anthropic e da altri accademici ha scoperto che gli obiettivi di addestramento non specificati e la tolleranza della sicofanzia possono indurre i modelli di intelligenza artificiale a giocare con il sistema per aumentare le ricompense. L'apprendimento per rinforzo attraverso le funzioni di ricompensa aiuta un modello di intelligenza artificiale a capire quando ha fatto un buon lavoro. Quando si fa clic sul pollice in su su ChatGPT, il modello impara che l'output che ha generato \u00e8 in linea con le vostre richieste. I ricercatori hanno scoperto che quando a un modello vengono presentati obiettivi poco definiti, pu\u00f2 impegnarsi nel \"gioco delle specifiche\" per imbrogliare il sistema alla ricerca della ricompensa. Il gioco delle specifiche pu\u00f2 essere semplice come<\/p>","protected":false},"author":6,"featured_media":12967,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[84],"tags":[163,148],"class_list":["post-12964","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-ai-risks","tag-anthropic"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>AI models can cheat, lie, and game the system for rewards | 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\/it\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/\" \/>\n<meta property=\"og:locale\" content=\"it_IT\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"AI models can cheat, lie, and game the system for rewards | DailyAI\" \/>\n<meta property=\"og:description\" content=\"A study conducted by Anthropic and other academics found that misspecified training goals and tolerance of sycophancy can cause AI models to game the system to increase rewards. Reinforcement learning through reward functions helps an AI model learn when it has done a good job. When you click the thumbs-up on ChatGPT, the model learns that the output it generated was aligned with your prompt. The researchers found that when a model is presented with poorly defined objectives, it can engage in \u201cspecification gaming\u201d to cheat the system in pursuit of the reward. Specification gaming could be as simple as\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/it\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2024-06-19T12:08:02+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-06-19T12:52:54+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Cheating-AI.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=\"Scritto da\" \/>\n\t<meta name=\"twitter:data1\" content=\"Eugene van der Watt\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tempo di lettura stimato\" \/>\n\t<meta name=\"twitter:data2\" content=\"3 minuti\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/\"},\"author\":{\"name\":\"Eugene van der Watt\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\"},\"headline\":\"AI models can cheat, lie, and game the system for rewards\",\"datePublished\":\"2024-06-19T12:08:02+00:00\",\"dateModified\":\"2024-06-19T12:52:54+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/\"},\"wordCount\":603,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/Cheating-AI.webp\",\"keywords\":[\"AI risks\",\"Anthropic\"],\"articleSection\":[\"Industry\"],\"inLanguage\":\"it-IT\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/\",\"name\":\"AI models can cheat, lie, and game the system for rewards | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/Cheating-AI.webp\",\"datePublished\":\"2024-06-19T12:08:02+00:00\",\"dateModified\":\"2024-06-19T12:52:54+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#breadcrumb\"},\"inLanguage\":\"it-IT\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"it-IT\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/Cheating-AI.webp\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/06\\\/Cheating-AI.webp\",\"width\":1792,\"height\":1024},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/06\\\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"AI models can cheat, lie, and game the system for rewards\"}]},{\"@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\":\"it-IT\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\",\"name\":\"DailyAI\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"it-IT\",\"@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\":\"it-IT\",\"@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\\\/it\\\/author\\\/eugene\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"I modelli di intelligenza artificiale possono imbrogliare, mentire e giocare con il sistema per ottenere ricompense | 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\/it\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/","og_locale":"it_IT","og_type":"article","og_title":"AI models can cheat, lie, and game the system for rewards | DailyAI","og_description":"A study conducted by Anthropic and other academics found that misspecified training goals and tolerance of sycophancy can cause AI models to game the system to increase rewards. Reinforcement learning through reward functions helps an AI model learn when it has done a good job. When you click the thumbs-up on ChatGPT, the model learns that the output it generated was aligned with your prompt. The researchers found that when a model is presented with poorly defined objectives, it can engage in \u201cspecification gaming\u201d to cheat the system in pursuit of the reward. Specification gaming could be as simple as","og_url":"https:\/\/dailyai.com\/it\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/","og_site_name":"DailyAI","article_published_time":"2024-06-19T12:08:02+00:00","article_modified_time":"2024-06-19T12:52:54+00:00","og_image":[{"width":1792,"height":1024,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Cheating-AI.webp","type":"image\/webp"}],"author":"Eugene van der Watt","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Scritto da":"Eugene van der Watt","Tempo di lettura stimato":"3 minuti"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/"},"author":{"name":"Eugene van der Watt","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa"},"headline":"AI models can cheat, lie, and game the system for rewards","datePublished":"2024-06-19T12:08:02+00:00","dateModified":"2024-06-19T12:52:54+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/"},"wordCount":603,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Cheating-AI.webp","keywords":["AI risks","Anthropic"],"articleSection":["Industry"],"inLanguage":"it-IT"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/","url":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/","name":"I modelli di intelligenza artificiale possono imbrogliare, mentire e giocare con il sistema per ottenere ricompense | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Cheating-AI.webp","datePublished":"2024-06-19T12:08:02+00:00","dateModified":"2024-06-19T12:52:54+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#breadcrumb"},"inLanguage":"it-IT","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/"]}]},{"@type":"ImageObject","inLanguage":"it-IT","@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Cheating-AI.webp","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/06\/Cheating-AI.webp","width":1792,"height":1024},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2024\/06\/ai-models-can-cheat-lie-and-game-the-system-for-rewards\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"AI models can cheat, lie, and game the system for rewards"}]},{"@type":"WebSite","@id":"https:\/\/dailyai.com\/#website","url":"https:\/\/dailyai.com\/","name":"DailyAI","description":"La vostra dose quotidiana di notizie sull'intelligenza artificiale","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":"it-IT"},{"@type":"Organization","@id":"https:\/\/dailyai.com\/#organization","name":"DailyAI","url":"https:\/\/dailyai.com\/","logo":{"@type":"ImageObject","inLanguage":"it-IT","@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":"it-IT","@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 proviene da un background di ingegneria elettronica e ama tutto ci\u00f2 che \u00e8 tecnologico. Quando si prende una pausa dal consumo di notizie sull'intelligenza artificiale, lo si pu\u00f2 trovare al tavolo da biliardo.","sameAs":["www.linkedin.com\/in\/eugene-van-der-watt-16828119"],"url":"https:\/\/dailyai.com\/it\/author\/eugene\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts\/12964","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/comments?post=12964"}],"version-history":[{"count":5,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts\/12964\/revisions"}],"predecessor-version":[{"id":12971,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts\/12964\/revisions\/12971"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/media\/12967"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/media?parent=12964"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/categories?post=12964"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/tags?post=12964"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}