rlm@302
|
1 Dear Professor Winston,
|
rlm@302
|
2
|
rlm@302
|
3 I'm ready for you to look through the work that I've done so far. It's
|
rlm@302
|
4 a sequence of posts describing the different simulated senses I've
|
rlm@302
|
5 implemented, with videos.
|
rlm@302
|
6
|
rlm@302
|
7 It's "blocks world reloaded", because like you say, you need multiple
|
rlm@302
|
8 senses to enable intelligence.
|
rlm@302
|
9
|
rlm@302
|
10 Please look through the videos and skim the text and tell me what
|
rlm@302
|
11 you think:
|
rlm@302
|
12
|
rlm@302
|
13 Introduction:
|
rlm@302
|
14 http://aurellem.org/cortex/html/intro.html
|
rlm@302
|
15 http://aurellem.org/cortex/html/sense.html
|
rlm@302
|
16
|
rlm@302
|
17 http://aurellem.org/cortex/html/body.html -- simulated physical bodies
|
rlm@302
|
18 http://aurellem.org/cortex/html/vision.html -- simulated eyes
|
rlm@302
|
19 http://aurellem.org/cortex/html/hearing.html -- simulated ears
|
rlm@302
|
20 http://aurellem.org/cortex/html/touch.html -- simulated skin/hairs
|
rlm@302
|
21 http://aurellem.org/cortex/html/proprioception.html -- simulated proprioception
|
rlm@302
|
22 http://aurellem.org/cortex/html/movement.html -- simulated muscles
|
rlm@302
|
23 http://aurellem.org/cortex/html/integration.html -- full demonstration
|
rlm@302
|
24
|
rlm@302
|
25 In particular, look at the video at
|
rlm@302
|
26 http://aurellem.org/cortex/html/integration.html. It shows a
|
rlm@302
|
27 simulated hand equipped with all of the senses I've built so far.
|
rlm@302
|
28
|
rlm@302
|
29 There's some more background information and full source code at
|
rlm@302
|
30 http://aurellem.org
|
rlm@302
|
31
|
rlm@302
|
32 If you can't see a video, let me know and I'll upload it to YouTube so
|
rlm@302
|
33 you can see it.
|
rlm@302
|
34
|
rlm@302
|
35
|
rlm@302
|
36
|
rlm@302
|
37
|
rlm@302
|
38 Now, I need your help moving forward. Can I use this work as a base
|
rlm@302
|
39 for a Masters thesis with you when I come back to MIT this coming Fall?
|
rlm@302
|
40 What critiques and project ideas do you have after looking through
|
rlm@302
|
41 what I've done so far?
|
rlm@302
|
42
|
rlm@302
|
43 I have some ideas on where I can go with this project but I think you
|
rlm@302
|
44 will have some better ones.
|
rlm@302
|
45
|
rlm@302
|
46 Here are some possible projects I might do with this as a base that I
|
rlm@302
|
47 think would be worthy Masters projects.
|
rlm@302
|
48
|
rlm@302
|
49 - HACKER for writing muscle-control programs : Presented with
|
rlm@302
|
50 low-level muscle control/ sense API, generate higher level programs
|
rlm@302
|
51 for accomplishing various stated goals. Example goals might be
|
rlm@302
|
52 "extend all your fingers" or "move your hand into the area with
|
rlm@302
|
53 blue light" or "decrease the angle of this joint". It would be
|
rlm@302
|
54 like Sussman's HACKER, except it would operate with much more data
|
rlm@302
|
55 in a more realistic world. Start off with "calisthenics" to
|
rlm@302
|
56 develop subroutines over the motor control API. This would be the
|
rlm@302
|
57 "spinal chord" of a more intelligent creature.
|
rlm@302
|
58
|
rlm@302
|
59 - Create hundreds of small creatures and have them do simple
|
rlm@302
|
60 simulated swarm tasks.
|
rlm@302
|
61
|
rlm@302
|
62 - A joint that learns what sort of joint it (cone, point-to-point,
|
rlm@302
|
63 hinge, etc.) is by correlating exploratory muscle movements with
|
rlm@302
|
64 vision.
|
rlm@302
|
65
|
rlm@302
|
66 - Something with cross-modal clustering using the rich sense
|
rlm@302
|
67 data. This might prove difficult due to the higher dimensionality
|
rlm@302
|
68 of my senses.
|
rlm@302
|
69
|
rlm@302
|
70 - Simulated Imagination --- this would involve a creature with an
|
rlm@302
|
71 effector which creates an /entire new sub-simulation/ where the
|
rlm@302
|
72 creature has direct control over placement/creation of objects via
|
rlm@302
|
73 simulated telekinesis. The creature observes this sub-world through
|
rlm@302
|
74 it's normal senses and uses its observations to make predictions
|
rlm@302
|
75 about it's top level world.
|
rlm@302
|
76
|
rlm@302
|
77 - Hook it up with Genesis --- I could make a "semantic effector"
|
rlm@302
|
78 which marks objects/sensory states with semantic information. In
|
rlm@302
|
79 conjunction with Simulated Imagination, and HACKER-like motor
|
rlm@302
|
80 control, Genesis might be able to ask simple questions like "stack
|
rlm@302
|
81 two blocks together and hit the top one with your hand; does the
|
rlm@302
|
82 bottom block move?" and the system could answer "yes". This would
|
rlm@302
|
83 be rather complicated to do and involves many of the above
|
rlm@302
|
84 projects, but there may be a way to scale it down to Master's
|
rlm@302
|
85 thesis size.
|
rlm@302
|
86
|
rlm@302
|
87 - Make a virtual computer in the virtual world which with which the
|
rlm@302
|
88 creature interacts using its fingers to press keys on a virtual
|
rlm@302
|
89 keyboard. The creature can access the internet, watch videos, take
|
rlm@302
|
90 over the world, anything it wants. (This is probably not worthy of
|
rlm@302
|
91 a Masters project, I just thought it was a neat idea. It's possible
|
rlm@302
|
92 to use videos/etc in the simulated world at any rate.)
|
rlm@302
|
93
|
rlm@302
|
94
|
rlm@302
|
95 I can't wait to hear your critiques and ideas. If you think I
|
rlm@302
|
96 shouldn't use this system as a base and should instead do something
|
rlm@302
|
97 else, that's fine too.
|
rlm@302
|
98
|
rlm@302
|
99 On a related note, can I be considered for the position of TA next
|
rlm@302
|
100 year for 6.034 or 6.xxx?
|
rlm@302
|
101
|
rlm@302
|
102 sincerely,
|
rlm@302
|
103 --Robert McIntyre
|