1.1 --- a/winston-intro.txt	Sat Feb 18 02:07:40 2012 -0600
     1.2 +++ b/winston-intro.txt	Sat Feb 18 10:28:14 2012 -0700
     1.3 @@ -1,103 +1,85 @@
     1.4  Dear Professor Winston,
     1.5  
     1.6 -I'm ready for you to look through the work that I've done so far. It's
     1.7 -a sequence of posts describing the different simulated senses I've
     1.8 -implemented, with videos. 
     1.9 +I've finished the first part of my project, building a framework for
    1.10 +virtual sensate creatures; I would like your help evaluating what I've
    1.11 +done so far, and deciding what to do next.
    1.12  
    1.13 -It's "blocks world reloaded", because like you say, you need multiple
    1.14 -senses to enable intelligence.
    1.15 +For the work I've done so far, I compiled the results into short
    1.16 +articles that explain how I implemented each sense, with videos that
    1.17 +show each sense in action. Please look through the articles, in
    1.18 +particular the video showcase, and tell me what you think.
    1.19  
    1.20 -Please look through the videos and skim the text and tell me what
    1.21 -you think:
    1.22 +Video Showcase : http://aurellem.localhost/cortex/org/cover.html
    1.23  
    1.24  Introduction:
    1.25  http://aurellem.org/cortex/html/intro.html
    1.26  http://aurellem.org/cortex/html/sense.html
    1.27  
    1.28 -http://aurellem.org/cortex/html/body.html  -- simulated physical bodies
    1.29 -http://aurellem.org/cortex/html/vision.html -- simulated eyes
    1.30 -http://aurellem.org/cortex/html/hearing.html -- simulated ears
    1.31 -http://aurellem.org/cortex/html/touch.html   -- simulated skin/hairs
    1.32 -http://aurellem.org/cortex/html/proprioception.html -- simulated proprioception
    1.33 -http://aurellem.org/cortex/html/movement.html       -- simulated muscles
    1.34 -http://aurellem.org/cortex/html/integration.html  -- full demonstration 
    1.35 +Physical Bodies : http://aurellem.org/cortex/html/body.html
    1.36 +Vision  : http://aurellem.org/cortex/html/vision.html
    1.37 +Hearing : http://aurellem.org/cortex/html/hearing.html
    1.38 +Touch   : http://aurellem.org/cortex/html/touch.html
    1.39 +Proprioception     : http://aurellem.org/cortex/html/proprioception.html
    1.40 +Muscles            : http://aurellem.org/cortex/html/movement.html
    1.41 +Full Demonstration : http://aurellem.org/cortex/html/integration.html
    1.42  
    1.43 -In particular, look at the video at
    1.44 -http://aurellem.org/cortex/html/integration.html.  It shows a
    1.45 -simulated hand equipped with all of the senses I've built so far.
    1.46 +I think this work could be a fruitful foundation for a Master's
    1.47 +thesis, so in particular, I'd like critiques, suggestions, and project
    1.48 +ideas. For example, here are some projects I think would be worthy, in
    1.49 +increasing order of complexity:
    1.50  
    1.51 -There's some more background information and full source code at 
    1.52 -http://aurellem.org
    1.53 +    * Create a self-powered joint that can determine its range of
    1.54 +      motion and joint type (hinge, cone, point-to-point, etc.) by
    1.55 +      making exploratory muscle movements and observing their effect.
    1.56  
    1.57 -If you can't see a video, let me know and I'll upload it to YouTube so
    1.58 -you can see it.
    1.59 +    * Develop an agent that writes and debugs low-level motor control
    1.60 +      programs to achieve simple goals like "look at the light" or
    1.61 +      "extend all of your fingers". These simple "calisthenic"
    1.62 +      programs could then be combined to form more elaborate
    1.63 +      procedures of motion, which in turn could be the basic
    1.64 +      instinctive reflexes in the "spinal cord" of some more advanced
    1.65 +      creature. (like Sussman's HACKER program but in a richer world)
    1.66  
    1.67 +    * Program a group of creatures that cooperate with each
    1.68 +      other. Because the creatures would be simulated, I could
    1.69 +      investigate computationally complex rules of behavior which
    1.70 +      still, from the group's point of view, would happen in "real
    1.71 +      time". Interactions could be as simple as cellular organisms
    1.72 +      communicating via flashing lights, or as complex as humanoids
    1.73 +      completing social tasks, etc.
    1.74  
    1.75 +    * Simulated Imagination -- this would involve a creature with an
    1.76 +      effector which creates an entire new sub-simulation where the
    1.77 +      creature has direct control over placement/creation of objects
    1.78 +      via simulated telekinesis. The creature observes this sub-world
    1.79 +      through it's normal senses and uses its observations to make
    1.80 +      predictions about it's top level world.
    1.81  
    1.82 +    * Integrate the simulated world with Genesis, so that Genesis
    1.83 +      could use the simulated world to answer questions about a
    1.84 +      proposed physical scenario. For example "You stack two blocks
    1.85 +      together, then hit the bottom block with your hand. Does the top
    1.86 +      block move?". This project is complicated and very large in
    1.87 +      scope, but it could be narrowed to focus on a single key
    1.88 +      aspect. For example, one key aspect of turning a scenario into a
    1.89 +      simulation is knowing when you're constructing "typical" or
    1.90 +      "atypical" examples of the scenario. So, a narrower project
    1.91 +      might simply learn about the edge cases of different scenarios
    1.92 +      (e.g. "A block stacked on top of another block is usually
    1.93 +      stable, provided the bottom block is large enough, and is not
    1.94 +      moving, and is level, etc."). With this knowledge, this kind of
    1.95 +      program could aid Genesis not only in answering common-sense
    1.96 +      questions, but in refining them: "A block is stacked on top of
    1.97 +      another block. Is it stable?"; "Usually, but do you know if the
    1.98 +      bottom block is slanted?", etc.
    1.99  
   1.100 -Now, I need your help moving forward. Can I use this work as a base
   1.101 -for a Masters thesis with you when I come back to MIT this coming Fall?
   1.102 -What critiques and project ideas do you have after looking through
   1.103 -what I've done so far?
   1.104 +These are some ideas, but I think you can come up with better ones. I
   1.105 +can't wait to hear your critiques and suggestions.
   1.106  
   1.107 -I have some ideas on where I can go with this project but I think you
   1.108 -will have some better ones.
   1.109 +Finally, regarding next year at MIT, can I be considered for the
   1.110 +position of TA for 6.034 or 6.xxx? Also, do you want me to return Fall
   1.111 +or Summer?
   1.112  
   1.113 -Here are some possible projects I might do with this as a base that I
   1.114 -think would be worthy Masters projects.
   1.115 +Sincerely, 
   1.116  
   1.117 - - HACKER for writing muscle-control programs : Presented with
   1.118 -   low-level muscle control/ sense API, generate higher level programs
   1.119 -   for accomplishing various stated goals. Example goals might be
   1.120 -   "extend all your fingers" or "move your hand into the area with
   1.121 -   blue light" or "decrease the angle of this joint".  It would be
   1.122 -   like Sussman's HACKER, except it would operate with much more data
   1.123 -   in a more realistic world.  Start off with "calisthenics" to
   1.124 -   develop subroutines over the motor control API.  This would be the
   1.125 -   "spinal chord" of a more intelligent creature.
   1.126 -
   1.127 - - Create hundreds of small creatures and have them do simple
   1.128 -   simulated swarm tasks.
   1.129 - 
   1.130 - - A joint that learns what sort of joint it (cone, point-to-point,
   1.131 -   hinge, etc.) is by correlating exploratory muscle movements with
   1.132 -   vision.
   1.133 -
   1.134 - - Something with cross-modal clustering using the rich sense
   1.135 -   data. This might prove difficult due to the higher dimensionality
   1.136 -   of my senses.
   1.137 -
   1.138 - - Simulated Imagination --- this would involve a creature with an
   1.139 -   effector which creates an /entire new sub-simulation/ where the
   1.140 -   creature has direct control over placement/creation of objects via
   1.141 -   simulated telekinesis. The creature observes this sub-world through
   1.142 -   it's normal senses and uses its observations to make predictions
   1.143 -   about it's top level world.
   1.144 -
   1.145 - - Hook it up with Genesis --- I could make a "semantic effector"
   1.146 -   which marks objects/sensory states with semantic information. In
   1.147 -   conjunction with Simulated Imagination, and HACKER-like motor
   1.148 -   control, Genesis might be able to ask simple questions like "stack
   1.149 -   two blocks together and hit the top one with your hand; does the
   1.150 -   bottom block move?" and the system could answer "yes".  This would
   1.151 -   be rather complicated to do and involves many of the above
   1.152 -   projects, but there may be a way to scale it down to Master's
   1.153 -   thesis size.
   1.154 -
   1.155 - - Make a virtual computer in the virtual world which with which the
   1.156 -   creature interacts using its fingers to press keys on a virtual
   1.157 -   keyboard. The creature can access the internet, watch videos, take
   1.158 -   over the world, anything it wants. (This is probably not worthy of
   1.159 -   a Masters project, I just thought it was a neat idea. It's possible
   1.160 -   to use videos/etc in the simulated world at any rate.)
   1.161 -
   1.162 -
   1.163 -I can't wait to hear your critiques and ideas.  If you think I
   1.164 -shouldn't use this system as a base and should instead do something
   1.165 -else, that's fine too. 
   1.166 -
   1.167 -On a related note, can I be considered for the position of TA next
   1.168 -year for 6.034 or 6.xxx?
   1.169 -
   1.170 -sincerely,
   1.171  --Robert McIntyre