{"id":337254,"date":"2017-03-31T04:52:37","date_gmt":"2017-03-30T20:52:37","guid":{"rendered":"http:\/\/techcrunch.com\/?p=1471153"},"modified":"2017-03-31T04:52:37","modified_gmt":"2017-03-30T20:52:37","slug":"baidus-ai-team-taught-a-virtual-agent-just-like-a-human-would-their-baby","status":"publish","type":"post","link":"https:\/\/people.utm.my\/asmawisham\/baidus-ai-team-taught-a-virtual-agent-just-like-a-human-would-their-baby\/","title":{"rendered":"Baidu\u2019s AI team taught a virtual agent just like a human would their baby"},"content":{"rendered":"

<\/p>\n

Baidu\u2019s artificial intelligence research team has achieved a significant milestone<\/a>: teaching a virtual agent \u201cliving\u201d in a 2D environment how to navigate its world using natural language commands, by first teaching it language through positive and negative reinforcement. The especially exciting thing, according to the scientists, is that the agent ended up developing a \u201czero-shot learning ability,\u201d which essentially\u00a0means that the AI agent developed a basic sense of grammar.<\/p>\n

You probably don\u2019t remember it from personal experience because it happened when you were a baby, but this is basically how parents teach their kids when very young. You show them images, repeat words and, eventually, with enough positive reinforcement, the kid can associate those words with those images and voil\u00e0 \u2014 it knows the names of things.<\/p>\n

Baidu\u2019s big breakthrough, though, is that the agent within its system can apply to new situations commands it\u2019s learned \u2014 computers aren\u2019t great at taking knowledge acquired before and applying it to new things. Here\u2019s the explanation direct from Baidu\u2019s own research team of why and how its system is different:<\/p>\n

\n

Applying past knowledge to a new<\/span>\u00a0task\u00a0<\/span>is very easy for human<\/span>s<\/span>\u00a0but still difficult for current end-to-end learning machines. Although\u00a0<\/span>machines may\u00a0<\/span>know what a \u201cdragon fruit\u201d look<\/span>s<\/span>\u00a0like, they can\u2019t\u00a0<\/span>perform the task \u201c<\/span>cut<\/strong>\u00a0the dragon fruit\u00a0with<\/strong>\u00a0a knife<\/strong><\/span>\u201d unless they\u00a0<\/span>have been<\/span>\u00a0explicitly trained with the dataset\u00a0<\/span>containing this\u00a0<\/span>command.\u00a0<\/span>By contrast, o<\/span>ur agent demonstrated the ability to transfer what they\u00a0<\/span>knew<\/span>\u00a0about the visual appearance of\u00a0<\/span>a<\/span>\u00a0dragon fruit as well as the task of \u201ccut<\/strong>\u00a0X\u00a0with<\/strong>\u00a0a knife<\/strong>\u201d successfully, without explicitly being trained to perform \u201ccut the dragon fruit with a knife<\/em><\/span>.<\/em>\u201d<\/span><\/p>\n<\/blockquote>\n

The ability to generalize use of a skill used previously is no small feat for artificial intelligences. It\u2019s a model that shows systems can learn and apply retained knowledge in a manner similar to humans, even if only in the limited realm of a simplistic, video game-like 2D environment. Which is very unlike me, in fact: I\u2019ll never learn.<\/p>\n","protected":false},"excerpt":{"rendered":"

\"\" Baidu’s artificial intelligence research team has achieved a significant milestone: teaching a virtual agent “living” in a 2D environment how to navigate its world using natural language commands, by first teaching it language through positive and negative reinforcement. Read More<\/a><\/p>\n

\n<\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a>\n<\/div>\n

\"\"\/ <\/p>\n","protected":false},"author":5817,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[25],"tags":[59,66,26],"class_list":["post-337254","post","type-post","status-publish","format-standard","hentry","category-technology","tag-media","tag-techcrunch","tag-technology"],"_links":{"self":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/337254","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/users\/5817"}],"replies":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/comments?post=337254"}],"version-history":[{"count":0,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/337254\/revisions"}],"wp:attachment":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/media?parent=337254"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/categories?post=337254"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/tags?post=337254"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}