陽明大學心智哲學研究所 ─2016年秋季系列演講─
講題：機器能有意識嗎？Can Machines Have Consciousness?
主講：金井良太 (Ryota Kanai)
Araya Brain Imaging, Tokyo, Japan
（http://www.araya.org/eng）
Earth-Life Science Institute (ELSI), Tokyo Institute of Technology
時間：9月28日（三）下午3:00-5:00
地點：陽明大學圖資大樓839室 （台北市北投區立農街二段155號，捷運石牌站）
主辦單位：陽明大學 心智哲學研究所、腦科學研究所、神經科學研究所、台灣意識研究群
報名網址：https://goo.gl/forms/VKAnieWrQTGhC1NP2
詳細訊息：http://goo.gl/UVhiZq
聯絡信箱：consciousness.tw@gmail.com
---
Speaker: Ryota Kanai
Araya Brain Imaging, Tokyo, Japan
Earth-Life Science Institute (ELSI), Tokyo Institute of Technology
Date & Time: 3:00-5:00 pm, September 28th, 2016
Location: Room 839, Library and Information Building, Institute of Neuroscience,
NYMU
Organized by Institute of Philosophy of Mind and Cognition,
Institute of Brain Science, Institute of Neuroscience, NYMU, and
Consciousness Research Group, Taiwan
Registration: https://goo.gl/forms/VKAnieWrQTGhC1NP2
More information: http://goo.gl/UVhiZq
Contact: consciousness.tw@gmail.com
本場演講將以英語進行。
以近年十分熱門的訊息整合理論（Information Integration Theory, IIT）出發，
探討意識的本質及人工意識的可能性。
非常歡迎對意識研究及其相關議題感興趣的朋友們來參加！
詳細的演講綱要如下↓
---
Abstract:
The recent surge of interest in artificial intelligence raises the question of
whether machines will ever possess human-level consciousness. This question
entails three fundamental issues about consciousness. First, can non-biological
physical objects have the capacity to have consciousness? In other words, do we
need neurons to have consciousness or can computer programs be conscious?
Second, can we ever be able to understand the neural and computational
principles of consciousness and implement them into machines?
Third, can we ever confirm that an artificial system has consciousness despite
the suggestion from the Hard problem that we cannot even prove that other
humans are conscious? In this talk, we will discuss possible answers to these
questions from the perspectives of the integrated information theory (IIT) of
consciousness. Specifically, I will argue that the axiomatic nature and the
identity claim of IIT enable us to determine whether conscious experience
exists in artificial systems. This could be understood in analogy with the way
other scientific theories allow inferences about reality that cannot be
directly observed (e.g., the presence of water in distant planets). However,
this argument hinges upon whether we accept IIT as the truth, and empirical
scrutiny of the theory will be essential. To do that, we are currently working
on the computational framework to compute the integrated information Φ in
biological and artificial neural systems and examine how Φ is related to
functions.
Finally, this constructive approach to consciousness highlights the question of
what consciousness is for? Can machines with consciousness outperform
unconscious AI? If so, in what respect? While these remain difficult questions
at present, potential functional benefits of conscious computation will be
discussed.
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 1.160.0.53
※ 文章網址: https://www.ptt.cc/bbs/Cognitive/M.1473178082.A.617.html
※ 編輯: gyosou (120.126.80.244), 09/12/2016 10:39:39