{"id":20,"date":"2019-09-25T20:52:55","date_gmt":"2019-09-25T20:52:55","guid":{"rendered":"https:\/\/www.rokits.org\/?p=20"},"modified":"2021-12-26T21:46:50","modified_gmt":"2021-12-26T21:46:50","slug":"playing-hide-and-seek-with-ai-and-privacy","status":"publish","type":"post","link":"https:\/\/www.rokits.org\/playing-hide-and-seek-with-ai-and-privacy\/","title":{"rendered":"Playing Hide And Seek With AI And Privacy"},"content":{"rendered":"\n

An AI Learns To Play Hide And Seek, Humans Learn to Seek Bots, And Google Is Recording Us Again<\/strong><\/p>\n\n\n\n

OpenAI released their new project<\/a> on September 17th. It\u2019s an AI that played over 500 million games of hide and seek. It learned to develop complex strategies and use tools. It even found flaws in the environment the programs created.<\/p>\n\n\n\n

The researchers built an environment with various objects inside a limited virtual space. The characters were incentivized to hide or seek, and to explore. The video below explains.<\/p>\n\n\n\n

\n