[Bmi] New from the MIT Press: Cognitive Robotics, edited by Angelo Cangelosi and Minoru Asada

[Xie Ming (Assoc Prof)]

Dear All,

I would like to share the following truths that I know from my studies and research works:


1.       Any entity in the universe has one list of can-do tasks and one list of can-not-do tasks. “Whether a human being or a robot is task-specific or not” is a superficial issue which could be totally ignored. It is wasting time to see such issue surfaces out again and again.

2.       Biological development refers to the reproduction in biological manufacturing. Such reproduction does not contribute to the design of mind.

3.       Social development is a process of gaining or growing experiences. Such development do contribute to the growth of knowledge and skills inside a mind. However, such progress is rooted on learning which has a prerequisite of “being educable” as outlined in my book published in 2003 by World Scientific Publishing Co.

4.       The discoveries made thousand years ago have told people that brain is brain while mind is mind. There are many experiments which could prove these discoveries. Such truth leads me to make the following statement: brain develops while mind learns.

5.       Humanoid robot includes body, brain, and mind. AI should be part of robotics. Robotics will drive the future growth of AI and prevents the third wave in AI to occur. Now, we have already entered the era of AI 3.0 (i.e. third generation of AI, but not third wave).

Warmest Regards

Ming XIE
http://society.robotics.sg<http://society.robotics.sg/>
http://www.pap.org.sg/our-party
http://personal.ntu.edu.sg/mmxie
http://www.editorialmanager.com/ijhr


From: Juyang Weng <juyang.weng at gmail.com>
Sent: Monday, May 23, 2022 2:49 AM
To: Angelo Cangelosi <angelo.cangelosi at manchester.ac.uk>; Minoru Asada <asada at ams.eng.osaka-u.ac.jp>
Cc: Xie Ming (Assoc Prof) <mmxie at ntu.edu.sg>; weng at cse.msu.edu; giulio.sandini at iit.it; Judith Bullent <jbullent at mit.edu>; Ron Arkin <arkin at gatech.edu>; bmilist <bmi at lists.cse.msu.edu>; Post Connectionists <connectionists at mailman.srv.cs.cmu.edu>
Subject: Re: New from the MIT Press: Cognitive Robotics, edited by Angelo Cangelosi and Minoru Asada

Dear Angelo and Minoru,
I would like to congratulate on the publication of the book.
I browsed the book quickly.   I would like to give you some quick feedback:

(1) About developmental robots:  Missed task-nonspecificity which is the flagship concept of Autonomous Mental Development (AMD) published in Science 2001.   For example, the six principles in Fig. 3.3 do not include task-nonspecificity,  The principles in Fig. 3.3 are vague, not unique timewise.  They appear to have appeared many times before 2000 when computational development was discussed in Workshop on Development and Learning. For example, what was new with "Embodiment" for robotics at all?   Is there a robot that does not have a body?   One important concept is "grounded" learning as in the acronym GENISAMA!   Another unfortunate phenomena throughout the book is that I do not see any task-nonspecific emergent representations.  All the representations that I have seen are symbolic in nature.
(2) The material in the book does not seem to give hope to end the brittleness of robotics.  Chapter 23 does not provide any architectures nor algorithms for conscious learning.  Developmental Networks, DN-1, DN-2 and now DN-3 suggest that conscious learning is a necessary condition for animals to develop animal-like intelligence (and awareness of the external world and self).  From the following paper, robots seem to need conscious learning to end their high brittleness (e.g., igerant about whether it is doing any task well).   I have received  some strong resistance to conscious learning (e.g., not willing to even review), but this resistance to new science should fail soon like any resistances to new sciences.
(3) I looked through the book series that Ron Arkin edited.  I was in his office about an AMD book that I was writing.   The book series that Ran Arkin edits lacks task-nonspecificity and also conscious learning.  Maybe the series can improve soon if Ron is open minded.

J. Weng, "An Algorithmic Theory of Conscious Learning", in Proc. 2022 3rd International Conf. on Artificial Intelligence in Electronics Engineering, pp. 1-10, Bangkok, Thailand, Jan. 21-23, 2022. PDF file<http://www.cse.msu.edu/~weng/research/ConsciousLearning-AIEE22rvsd-cite.pdf>.

Best regards,
-John
---
Juyang (John) Weng
________________________________

CONFIDENTIALITY: This email is intended solely for the person(s) named and may be confidential and/or privileged. If you are not the intended recipient, please delete it, notify us and do not copy, use, or disclose its contents.
Towards a sustainable earth: Print only when necessary. Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cse.msu.edu/pipermail/bmi/attachments/20220523/b5c5f653/attachment-0001.html>