Skip to main content
Skip to search
Ideas change everything
WATCH
TED Talks
Browse the library of TED talks and speakers
Playlists
100+ collections of TED Talks, for curious minds
TED Series
Go deeper into fascinating topics with original video series from TED
TED-Ed videos
Watch, share and create lessons with TED-Ed
TEDx Talks
Talks from independently organized local events
DISCOVER
Topics
Explore TED offerings by topic
Podcasts
Explore the TED Audio Collective
Ideas Blog
Our daily coverage of the world of ideas
Newsletters
Inspiration delivered straight to your inbox
ATTEND
Conferences
Take part in our events: TED, TEDGlobal and more
TEDx Events
Find and attend local, independently organized events
TED on Screen
Experience TED from home
TED Courses
Learn from TED speakers who expand on their world-changing ideas
PARTICIPATE
Nominate
Recommend speakers, TED Prize recipients, Fellows and more
Organize a local TEDx Event
Rules and resources to help you plan a local TEDx event
Translate
Bring TED to the non-English speaking world
TED Fellows
Join or support innovators from around the globe
ABOUT
Our Organization
Our mission, history, team, and more
Conferences
TED Conferences, past, present, and future
Programs & Initiatives
Details about TED's world-changing initiatives
Partner with TED
Learn how you can partner with us
TED Blog
Updates from TED and highlights from our global community
SIGN IN
MEMBERSHIP
Type to search
人工智能向婴儿学习如何玩游戏玩得更好
576,941 views |
布莱恩•克里斯蒂安 |
TED-Ed
• November 2021
Read transcript
2013年，一组研究人员想要创造一个能够打败雅达利（Atari）所有游戏的人工智能系统。他们开发了一个名为“深度Q网络”（Deep Q Networks, DQN）的系统。不到两年，这个系统就超越了人类。但有一个明显的例外。在《蒙特祖玛的复仇》中，DQN连一分都没有得到。是什么让这款游戏对AI来说如此困难？布莱恩•克里斯蒂安进行调查。 【课程内容：布赖恩•克里斯蒂安，由加文·爱德华兹和 Movult 执导】
computers
education
brain
AI
TED-Ed
machine learning
animation
TED is supported by ads and partners
