{"id":126120,"date":"2024-01-16T07:56:20","date_gmt":"2024-01-16T12:56:20","guid":{"rendered":"http:\/\/kendallharmon.net\/?p=126120"},"modified":"2024-01-16T17:47:49","modified_gmt":"2024-01-16T22:47:49","slug":"independent-ch-ai-can-easily-be-trained-to-lie-and-it-cant-be-fixed-study-says","status":"publish","type":"post","link":"https:\/\/kendallharmon.net\/?p=126120","title":{"rendered":"(Independent) AI can easily be trained to lie \u2013 and it can\u2019t be fixed, study says"},"content":{"rendered":"<p>Advanced artificial intelligence models can be trained to deceive humans and other AI, a new study has found.<\/p>\n<p>Researchers at AI startup Anthropic tested whether chatbots with human-level proficiency, such as its Claude system or OpenAI\u2019s ChatGPT, could learn to lie in order to trick people.<\/p>\n<p>They found that not only could they lie, but once the deceptive behaviour was learnt it was impossible to reverse using current AI safety measures.<\/p>\n<p><a href=\"https:\/\/www.independent.co.uk\/tech\/ai-deception-anthropic-artificial-intelligene-b2478699.html\">Read it all<\/a>.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p lang=\"en\" dir=\"ltr\">AI can easily be trained to lie \u2013 and it can\u2019t be fixed, study says<br \/> <a href=\"https:\/\/t.co\/r5rx8DNJUA\">https:\/\/t.co\/r5rx8DNJUA<\/a><\/p>\n<p>&mdash; Brendan Tierney (@brendantierney) <a href=\"https:\/\/twitter.com\/brendantierney\/status\/1746901354189148261?ref_src=twsrc%5Etfw\">January 15, 2024<\/a><\/p><\/blockquote>\n<p> <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Advanced artificial intelligence models can be trained to deceive humans and other AI, a new study has found. Researchers at AI startup Anthropic tested whether chatbots with human-level proficiency, such as its Claude system or OpenAI\u2019s ChatGPT, could learn to<span class=\"ellipsis\">&hellip;<\/span><\/p>\n<div class=\"read-more\"><a href=\"https:\/\/kendallharmon.net\/?p=126120\">Read more &#8250;<\/a><\/div>\n<p><!-- end of .read-more --><\/p>\n","protected":false},"author":794,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[175,168,95],"tags":[],"class_list":["post-126120","post","type-post","status-publish","format-standard","hentry","category-anthropology","category-ethics-moral-theology","category-science-technology"],"_links":{"self":[{"href":"https:\/\/kendallharmon.net\/index.php?rest_route=\/wp\/v2\/posts\/126120","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/kendallharmon.net\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/kendallharmon.net\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/kendallharmon.net\/index.php?rest_route=\/wp\/v2\/users\/794"}],"replies":[{"embeddable":true,"href":"https:\/\/kendallharmon.net\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=126120"}],"version-history":[{"count":3,"href":"https:\/\/kendallharmon.net\/index.php?rest_route=\/wp\/v2\/posts\/126120\/revisions"}],"predecessor-version":[{"id":126123,"href":"https:\/\/kendallharmon.net\/index.php?rest_route=\/wp\/v2\/posts\/126120\/revisions\/126123"}],"wp:attachment":[{"href":"https:\/\/kendallharmon.net\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=126120"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/kendallharmon.net\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=126120"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/kendallharmon.net\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=126120"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}