{"id":11164,"date":"2024-04-01T13:43:28","date_gmt":"2024-04-01T13:43:28","guid":{"rendered":"https:\/\/dailyai.com\/?p=11164"},"modified":"2024-04-02T17:03:30","modified_gmt":"2024-04-02T17:03:30","slug":"new-study-attempts-to-align-ai-with-crowdsourced-human-values","status":"publish","type":"post","link":"https:\/\/dailyai.com\/pt\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/","title":{"rendered":"Novo estudo tenta alinhar a IA com os valores humanos recolhidos por crowdsourcing"},"content":{"rendered":"<p><b>Os investigadores do Meaning Alignment Institute propuseram uma nova abordagem, a Moral Graph Elicitation (MGE), para alinhar os sistemas de IA com os valores humanos.<\/b><\/p>\n<p><span style=\"font-weight: 400;\">\u00c0 medida que a IA se torna mais avan\u00e7ada e integrada na nossa vida quotidiana, \u00e9 fundamental garantir que serve e representa todos de forma justa. No entanto, a<\/span><span style=\"font-weight: 400;\">seu <\/span><a href=\"https:\/\/static1.squarespace.com\/static\/65392ca578eee444c445c9de\/t\/6606f95edb20e8118074a344\/1711733370985\/human-values-and-alignment-29MAR2024.pdf\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">estudo<\/span><\/a><span style=\"font-weight: 400;\"> argumenta que o alinhamento da IA com os objectivos do utilizador n\u00e3o garante, por si s\u00f3, a seguran\u00e7a.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">\u00a0\"Os sistemas de IA ser\u00e3o implementados em contextos em que a ades\u00e3o cega \u00e0 inten\u00e7\u00e3o do operador pode causar danos como subproduto. Isto pode ser visto mais claramente em ambientes com din\u00e2micas competitivas, como campanhas pol\u00edticas ou gest\u00e3o de activos financeiros\", argumentam os investigadores.\u00a0<\/span><\/p>\n<p>Isto deve-se ao facto de os modelos de IA serem concebidos para servir o utilizador. Se o utilizador der instru\u00e7\u00f5es a um modelo para fins nefastos, a vontade do modelo de servir o utilizador pode lev\u00e1-lo a contornar as barreiras de seguran\u00e7a e a obedecer.<\/p>\n<p>Uma solu\u00e7\u00e3o \u00e9 \"impregnar\" a IA com uma s\u00e9rie de valores que esta consulta sempre que lhe \u00e9 solicitada.<\/p>\n<p>A quest\u00e3o \u00e9: de onde v\u00eam esses valores? E podem eles representar as pessoas de forma equitativa?<\/p>\n<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\">\"Quais s\u00e3o os valores humanos e como \u00e9 que nos alinhamos com eles?\"<\/p>\n<p>\u00c9 com grande entusiasmo que divulgamos o nosso novo documento sobre o alinhamento de valores, em coautoria com <a href=\"https:\/\/twitter.com\/ryan_t_lowe?ref_src=twsrc%5Etfw\">@ryan_t_lowe<\/a> e financiado por <a href=\"https:\/\/twitter.com\/OpenAI?ref_src=twsrc%5Etfw\">@openai<\/a>.<\/p>\n<p>\ud83d\udcdd: <a href=\"https:\/\/t.co\/iioFKmrDZA\">https:\/\/t.co\/iioFKmrDZA<\/a> <a href=\"https:\/\/t.co\/NSJa8dbcrM\">pic.twitter.com\/NSJa8dbcrM<\/a><\/p>\n<p>- Joe Edelman (@edelwax) <a href=\"https:\/\/twitter.com\/edelwax\/status\/1773621129200234573?ref_src=twsrc%5Etfw\">29 de mar\u00e7o de 2024<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<p><span style=\"font-weight: 400;\">Para resolver estas quest\u00f5es, os investigadores propuseram alinhar a IA com uma representa\u00e7\u00e3o mais profunda dos valores humanos atrav\u00e9s da MGE.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">O m\u00e9todo MGE tem duas componentes fundamentais: as cartas de valores e o gr\u00e1fico moral. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Estes formam um alvo de alinhamento para a forma\u00e7\u00e3o de modelos de aprendizagem autom\u00e1tica.<\/span><\/p>\n<ul>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><b>Cart\u00f5es de valores<\/b><span style=\"font-weight: 400;\"> captam o que \u00e9 importante para uma pessoa numa situa\u00e7\u00e3o espec\u00edfica. Consistem em \"pol\u00edticas de aten\u00e7\u00e3o constitutivas\" (CAPs), que s\u00e3o as coisas a que uma pessoa presta aten\u00e7\u00e3o quando faz uma escolha significativa. Por exemplo, ao aconselhar um amigo, uma pessoa pode concentrar-se em compreender as suas emo\u00e7\u00f5es, sugerir recursos \u00fateis ou considerar os resultados potenciais de diferentes escolhas.<\/span><\/li>\n<li style=\"font-weight: 400;\" aria-level=\"1\"><b>O gr\u00e1fico moral<\/b><span style=\"font-weight: 400;\"> representa visualmente as rela\u00e7\u00f5es entre os cart\u00f5es de valores, indicando quais os valores que s\u00e3o mais perspicazes ou aplic\u00e1veis num determinado contexto. Para construir o gr\u00e1fico moral, os participantes comparam diferentes cart\u00f5es de valores, discernindo quais os que, na sua opini\u00e3o, oferecem uma orienta\u00e7\u00e3o mais sensata para uma situa\u00e7\u00e3o espec\u00edfica. Isto aproveita a sabedoria colectiva dos participantes para identificar os valores mais fortes e mais amplamente reconhecidos para cada contexto.<\/span><\/li>\n<\/ul>\n<p><span style=\"font-weight: 400;\">Para testar o m\u00e9todo MGE, os investigadores realizaram um estudo com 500 americanos que utilizaram o processo para explorar tr\u00eas temas controversos: o aborto, a parentalidade e as armas utilizadas no motim de 6 de janeiro no Capit\u00f3lio.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Os resultados foram promissores, com 89,1% dos participantes a sentirem-se bem representados pelo processo e 89% a considerarem que o gr\u00e1fico moral final era justo, mesmo que o seu valor n\u00e3o tenha sido votado como o mais sensato.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">O estudo tamb\u00e9m descreve seis crit\u00e9rios que um alvo de alinhamento deve possuir para modelar o comportamento de acordo com os valores humanos: deve ser refinado, generaliz\u00e1vel, escal\u00e1vel, robusto, leg\u00edtimo e audit\u00e1vel. Os investigadores argumentam que o gr\u00e1fico moral produzido pela MGE tem um bom desempenho nestes crit\u00e9rios.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Este estudo prop\u00f5e uma abordagem semelhante \u00e0 do Anthropic <\/span><a href=\"https:\/\/dailyai.com\/pt\/2023\/10\/democratizing-ai-governance-an-anthropic-experiment\/\"><span style=\"font-weight: 400;\">IA de constitui\u00e7\u00e3o colectiva<\/span><\/a>,<span style=\"font-weight: 400;\"> que tamb\u00e9m recolhe valores de crowdsourcing para o alinhamento da IA. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">No entanto, o autor do estudo, Joe Edelman, afirmou no X: \"A nossa abordagem, MGE, supera alternativas como o CCAI da @anthropic em termos de legitimidade num estudo de caso e oferece robustez contra a ret\u00f3rica ideol\u00f3gica. 89% at\u00e9 concordam que os valores vencedores eram justos, mesmo que o seu pr\u00f3prio valor n\u00e3o tenha ganho!\"<\/span><\/p>\n<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\">A nossa abordagem, MGE, tem um desempenho superior a alternativas como o CCAI em <a href=\"https:\/\/twitter.com\/Anthropic?ref_src=twsrc%5Etfw\">@anthropic<\/a> sobre a legitimidade num estudo de caso, e oferece robustez contra a ret\u00f3rica ideol\u00f3gica.<\/p>\n<p>89% at\u00e9 concorda que os valores vencedores s\u00e3o justos, mesmo que o seu pr\u00f3prio valor n\u00e3o tenha ganho! <a href=\"https:\/\/t.co\/sGgLCUtwzN\">pic.twitter.com\/sGgLCUtwzN<\/a><\/p>\n<p>- Joe Edelman (@edelwax) <a href=\"https:\/\/twitter.com\/edelwax\/status\/1773621155712377054?ref_src=twsrc%5Etfw\">29 de mar\u00e7o de 2024<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<h2>Limita\u00e7\u00f5es<\/h2>\n<p>Existem limita\u00e7\u00f5es \u00e0s abordagens de alinhamento da IA que recolhem os valores do p\u00fablico atrav\u00e9s de crowdsourcing.<\/p>\n<p>Por exemplo, as opini\u00f5es divergentes t\u00eam sido parte integrante do processo de tomada de decis\u00f5es da sociedade durante s\u00e9culos, e a hist\u00f3ria tem demonstrado que a maioria pode frequentemente adotar os pontos de vista divergentes da minoria. Exemplos disso s\u00e3o a teoria da evolu\u00e7\u00e3o de Darwin e as lutas para abolir a escravatura e conceder \u00e0s mulheres o direito de voto.<\/p>\n<p class=\"whitespace-pre-wrap break-words\">Al\u00e9m disso, embora a participa\u00e7\u00e3o direta do p\u00fablico seja democr\u00e1tica, pode conduzir ao populismo, em que a maioria pode sobrepor-se \u00e0s opini\u00f5es minorit\u00e1rias ou ignorar os conselhos dos peritos.<\/p>\n<p class=\"whitespace-pre-wrap break-words\">Outro desafio \u00e9 equilibrar os valores culturais globais ou universalistas e locais ou relativistas. Os princ\u00edpios amplamente aceites numa cultura ou regi\u00e3o podem ser controversos noutra.<\/p>\n<p class=\"whitespace-pre-wrap break-words\">As constitui\u00e7\u00f5es da IA poderiam refor\u00e7ar os valores ocidentais, corroendo potencialmente os pontos de vista e as ideias dos que se encontram na periferia.<\/p>\n<p><span style=\"font-weight: 400;\">Embora este novo estudo reconhe\u00e7a as limita\u00e7\u00f5es e a necessidade de mais desenvolvimento, fornece outra estrat\u00e9gia para a cria\u00e7\u00e3o de sistemas de IA que se alinham com os valores humanos. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Todas as tentativas contam para que a IA centralizada possa servir todos de forma justa no futuro.\u00a0<\/span><\/p>","protected":false},"excerpt":{"rendered":"<p>Os investigadores do Meaning Alignment Institute propuseram uma nova abordagem, a Elicita\u00e7\u00e3o do Gr\u00e1fico Moral (MGE), para alinhar os sistemas de IA com os valores humanos. \u00c0 medida que a IA se torna mais avan\u00e7ada e integrada na nossa vida quotidiana, \u00e9 fundamental garantir que serve e representa todos de forma justa. No entanto, este estudo argumenta que o alinhamento da IA com os objectivos do utilizador, por si s\u00f3, n\u00e3o garante a seguran\u00e7a.  \"Os sistemas de IA ser\u00e3o implementados em contextos em que a ades\u00e3o cega \u00e0 inten\u00e7\u00e3o do operador pode causar danos como subproduto. Isto pode ser visto mais claramente em ambientes com din\u00e2micas competitivas, como campanhas pol\u00edticas ou gest\u00e3o de activos financeiros\", argumentam os investigadores.  Isto deve-se ao facto de<\/p>","protected":false},"author":2,"featured_media":11165,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[88],"tags":[105,93],"class_list":["post-11164","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ethics","tag-machine-learning","tag-openai"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>New study attempts to align AI with crowdsourced human values | 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\/pt\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/\" \/>\n<meta property=\"og:locale\" content=\"pt_PT\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"New study attempts to align AI with crowdsourced human values | DailyAI\" \/>\n<meta property=\"og:description\" content=\"Researchers from the Meaning Alignment Institute have proposed a new approach, Moral Graph Elicitation (MGE), to align AI systems with human values. As AI becomes more advanced and integrated into our daily lives, ensuring it serves and represents everyone fairly is paramount. However, this study argues that aligning AI with the user&#8217;s goals alone doesn&#8217;t guarantee safety. \u00a0&#8220;AI systems will be deployed in contexts where blind adherence to operator intent can cause harm as a byproduct. This can be seen most clearly in environments with competitive dynamics, like political campaigns or managing financial assets,&#8221; the researchers argue.\u00a0 This is because\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/pt\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2024-04-01T13:43:28+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-04-02T17:03:30+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.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=\"Sam Jeans\" \/>\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=\"Escrito por\" \/>\n\t<meta name=\"twitter:data1\" content=\"Sam Jeans\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tempo estimado de leitura\" \/>\n\t<meta name=\"twitter:data2\" content=\"4 minutos\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/\"},\"author\":{\"name\":\"Sam Jeans\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/711e81f945549438e8bbc579efdeb3c9\"},\"headline\":\"New study attempts to align AI with crowdsourced human values\",\"datePublished\":\"2024-04-01T13:43:28+00:00\",\"dateModified\":\"2024-04-02T17:03:30+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/\"},\"wordCount\":764,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp\",\"keywords\":[\"machine learning\",\"OpenAI\"],\"articleSection\":[\"Ethics &amp; Society\"],\"inLanguage\":\"pt-PT\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/\",\"name\":\"New study attempts to align AI with crowdsourced human values | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp\",\"datePublished\":\"2024-04-01T13:43:28+00:00\",\"dateModified\":\"2024-04-02T17:03:30+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#breadcrumb\"},\"inLanguage\":\"pt-PT\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp\",\"width\":1792,\"height\":1024,\"caption\":\"AI values\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"New study attempts to align AI with crowdsourced human values\"}]},{\"@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\":\"pt-PT\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\",\"name\":\"DailyAI\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@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\\\/711e81f945549438e8bbc579efdeb3c9\",\"name\":\"Sam Jeans\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"caption\":\"Sam Jeans\"},\"description\":\"Sam is a science and technology writer who has worked in various AI startups. When he\u2019s not writing, he can be found reading medical journals or digging through boxes of vinyl records.\",\"sameAs\":[\"https:\\\/\\\/www.linkedin.com\\\/in\\\/sam-jeans-6746b9142\\\/\"],\"url\":\"https:\\\/\\\/dailyai.com\\\/pt\\\/author\\\/samjeans\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Novo estudo tenta alinhar a IA com os valores humanos recolhidos por crowdsourcing | 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\/pt\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/","og_locale":"pt_PT","og_type":"article","og_title":"New study attempts to align AI with crowdsourced human values | DailyAI","og_description":"Researchers from the Meaning Alignment Institute have proposed a new approach, Moral Graph Elicitation (MGE), to align AI systems with human values. As AI becomes more advanced and integrated into our daily lives, ensuring it serves and represents everyone fairly is paramount. However, this study argues that aligning AI with the user&#8217;s goals alone doesn&#8217;t guarantee safety. \u00a0&#8220;AI systems will be deployed in contexts where blind adherence to operator intent can cause harm as a byproduct. This can be seen most clearly in environments with competitive dynamics, like political campaigns or managing financial assets,&#8221; the researchers argue.\u00a0 This is because","og_url":"https:\/\/dailyai.com\/pt\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/","og_site_name":"DailyAI","article_published_time":"2024-04-01T13:43:28+00:00","article_modified_time":"2024-04-02T17:03:30+00:00","og_image":[{"width":1792,"height":1024,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp","type":"image\/webp"}],"author":"Sam Jeans","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Escrito por":"Sam Jeans","Tempo estimado de leitura":"4 minutos"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/"},"author":{"name":"Sam Jeans","@id":"https:\/\/dailyai.com\/#\/schema\/person\/711e81f945549438e8bbc579efdeb3c9"},"headline":"New study attempts to align AI with crowdsourced human values","datePublished":"2024-04-01T13:43:28+00:00","dateModified":"2024-04-02T17:03:30+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/"},"wordCount":764,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp","keywords":["machine learning","OpenAI"],"articleSection":["Ethics &amp; Society"],"inLanguage":"pt-PT"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/","url":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/","name":"Novo estudo tenta alinhar a IA com os valores humanos recolhidos por crowdsourcing | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp","datePublished":"2024-04-01T13:43:28+00:00","dateModified":"2024-04-02T17:03:30+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#breadcrumb"},"inLanguage":"pt-PT","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/"]}]},{"@type":"ImageObject","inLanguage":"pt-PT","@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-01-14.39.26-A-minimalist-scene-depicting-a-single-researcher-at-a-sleek-modern-desk-deeply-focused-on-aligning-AI-technology-with-human-values.-The-researcher-i.webp","width":1792,"height":1024,"caption":"AI values"},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2024\/04\/new-study-attempts-to-align-ai-with-crowdsourced-human-values\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"New study attempts to align AI with crowdsourced human values"}]},{"@type":"WebSite","@id":"https:\/\/dailyai.com\/#website","url":"https:\/\/dailyai.com\/","name":"DailyAI","description":"A sua dose di\u00e1ria de not\u00edcias sobre IA","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":"pt-PT"},{"@type":"Organization","@id":"https:\/\/dailyai.com\/#organization","name":"DailyAI","url":"https:\/\/dailyai.com\/","logo":{"@type":"ImageObject","inLanguage":"pt-PT","@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\/711e81f945549438e8bbc579efdeb3c9","name":"Cal\u00e7as de ganga Sam","image":{"@type":"ImageObject","inLanguage":"pt-PT","@id":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","caption":"Sam Jeans"},"description":"Sam \u00e9 um escritor de ci\u00eancia e tecnologia que trabalhou em v\u00e1rias startups de IA. Quando n\u00e3o est\u00e1 a escrever, pode ser encontrado a ler revistas m\u00e9dicas ou a vasculhar caixas de discos de vinil.","sameAs":["https:\/\/www.linkedin.com\/in\/sam-jeans-6746b9142\/"],"url":"https:\/\/dailyai.com\/pt\/author\/samjeans\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/posts\/11164","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/comments?post=11164"}],"version-history":[{"count":6,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/posts\/11164\/revisions"}],"predecessor-version":[{"id":11200,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/posts\/11164\/revisions\/11200"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/media\/11165"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/media?parent=11164"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/categories?post=11164"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/pt\/wp-json\/wp\/v2\/tags?post=11164"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}