Instructor
|
Prof. Marvin Minsky
|
Teaching Assistants
|
Dylan Holmes (dxh@mit), Robert McIntyre (rlm@mit), Heikki Ruuska.
|
When and where
|
Wednesdays from 7-9pm, in
32-141
|
Office hours
|
, and by appointment.
|
Note (30 Apr) Prof. Minsky will be attending office hours as
time permits. Moreover, we may hold an additional session of office
hours on either Thu or Fri, depending on your preference.
About the course
The subject of this course is to discuss theories about how minds
work. We will emphasize aspects of thinking that are so poorly
understood that they are still considered to be more philosophical
than scientific. Our goal is to replace ill-defined folk theories of
‘consciousness’, ‘self’ and ‘emotion’ with more concrete computational concepts.
This course centers around two books, Society of Mind and The
Emotion Machine. Both books are available in paperback at the
Coop, and the full text of The Emotion Machine is also available on Prof. Minsky's
homepage.
There are no course prerequisites for 6.868. However, you are
expected to have read Society of Mind in advance.
Expectations (yours and ours)
This class takes the form of an interactive seminar, in which you
will be able to actively discuss and debate ideas. We expect that
you will be thoughtful, inquisitive, and prepared.
- You should come to every class prepared to ask a clear and
important question. This class presents a real opportunity to
figure out hard problems about consciousness, intelligence,
selfhood, and human nature—among others. To be successful,
you need only study the material carefully and critically outside of
class.
- You should read Society of Mind. Because this class
immediately builds upon the ideas presented in the book, you will benefit
the most if you have carefully read it ahead of time.
Now about your expectations:
- Lectures. The lectures will be extemporaneous;
at the start of class, you'll raise questions, criticisms,
and suggestions about the material,
or about your own theories, which will then guide
the course of Prof. Minsky's talks.
- Student-driven discussions. We will have student-driven
discussions in the latter half of most lectures. This will involve
small groups of students forming a panel in the front of the class
and talking among themselves and with the audience about a
particular topic— very much like a talk show.
You'll know the topic in advance, and
will know which week will be your turn, so you'll be able to
plan and prepare accordingly.
We will provide supplementary materials (articles, book
chapters, websites, online videos, etc.) and questions to help you
prepare for the student-driven discussion. We expect that the discussions
will be informal and fun— not intimidating — however,
if you believe that this will not be the case for you, you should
contact the TAs so that we can discuss it.
- Visiting lecturers. We anticipate a number of eminent guest speakers throughout the
semester. These speakers could give a talk about their specialty, or
engage in discussion with the audience about a pre-determined
paper. In short, you'll have the opportunity to hear and speak with
several luminaries of artifical intelligence and other nearby fields.
- Final project. The culmination of your work in the course
will be an open-ended final project in which you'll apply your understanding
of the ideas in 6.868. This project can take several
different forms, depending on your individual interests. More
details about the project follow:
- If your project involves programming, you should submit your
source code, along with a 3-5 page description detailing your
problem statement, documentation for your code, and conclusions from
your project.
- If your project is entirely written, it should be 10-15 pages in
length, and should include proper citations.
- For other kinds of projects, contact us so we can discuss
equivalent grading criteria. (In most cases, your requirements will
be analogous to those for programming projects.)
- Standard formatting guidelines apply: please use 11-12pt serif
font (for example, Times New Roman), single or 1.5 spacing, with
1″ margins. Your written components should be submitted in
either PDF or plain text format.
- In order for your project to be successful, you will need to
demonstrate your comprehension of ideas in Society of Mind. For
example, you should replace vague suitcase words
like thinking, understanding, knowing with more
precise and explicit terms. More fundamentally, you should use what
you learned from reading Society of Mind and The Emotion
Machine in your project— if you could have implemented
your project proposal using what you knew before reading
Prof. Minsky's ideas, it is at least doubtful that it would make a
good final project for this class.
-
We are happy to help you with your project. In addition to e-mailing
us with questions or attending office hours, you may
submit your project to the staff (society-of-mind@mit.edu)
for review at any time, and we will give you constructive
feedback. Your project is finally due by 5pm on the last day of
classes, 16 May. The official due date is Thursday, although we will accept papers until Friday.
- Succeeding in 6.868. Your success in this in this course will be determined by your
preparation and participation in class discussion, and by your
ability to understand and employ the main ideas of the class.
We measure your participation through your involvement in
the lectures and in the student-led discussion, as well as through
brief written assignments which we may collect from time to time. We
measure your ability to understand and employ the main ideas of the
class through your final project.
Other Resources
You can watch the video clips from the
interactive Society of Mind Voyager CD-ROM here.
You can read a summary of your responses to Chapter 6 here.
Course topics
Week |
Lecture |
Preparation |
Assignment |
1
Feb 6
|
Why it's important to understand intelligence
|
Read
the introduction
to The Emotion Machine— and, of course, bring
your questions to class.
|
Fill out the registration
survey. (The lottery has been concluded.)
|
2
Feb 13
|
Applications of common sense (Henry Lieberman)
|
Optional background reading: Beating
Common Sense into Interactive Applications and AnalogySpace: Reducing the Dimensionality of Commonsense Knowledge
|
-
|
3
Feb 20
|
EM 1: Falling in Love
|
Read Chapter 1
of The Emotion Machine.
|
Emotions and rationality
|
4
Feb 27
|
Preparing for intelligent machines (Ed Fredkin)
|
Read about the important issues,
and of course bring your questions and criticisms to class.
|
-
|
5
Mar 6
|
EM 4: Consciousness
|
Read Chapter 4
of The Emotion Machine— and, of course, bring your questions to class.
|
Unpacking consciousness
|
6
Mar 13
|
Learning how to learn (Mike Hawley)
|
Evaluate how learning how we learn can
empower us to improve and control our mental lives.
|
-
|
7
Mar 20
|
EM 5: Levels of Mental Activities
|
Read Chapter 5
of The Emotion Machine, and submit your questions to the staff.
|
Levels of Mental Activities
+ Project Proposal
|
Spring vacation
|
|
8
Apr 3
|
Designing common sense systems (Danny Hillis)
|
Optional background reading: Danny Hillis's thesis on the Connection
Machine, and The Clock
of the Long Now.
|
-
|
9
Apr 10
|
EM 6: Common Sense
|
Read Chapter 6
of The Emotion Machine, and submit your questions to the staff.
|
Common Sense
|
10
Apr 17
|
Implementing the Society of Mind (Dustin Smith & Heikki Ruuska)
|
Read Chapter 7
of The Emotion Machine, and submit your questions to the staff.
|
Implementing the Society
of Mind + Thinking
|
11
Apr 24
|
EM 7: Thinking
|
Read, rank, and review the class responses
to Chapter 6
and Chapter 7.
|
Read & Review
|