Community Archive

🧵 View Thread

🧵 Thread (7 tweets)

Placeholder
UltimApe@ultimape• over 5 years ago

I just realized why I am all about "embodied cognition"I read this when I was 7: https://t.co/xzGSQhwx6X

3 0
12/27/2019
Placeholder
UltimApe@ultimape• over 5 years ago
Replying to @ultimape

I was ~15 when I read this:"Theory of Mind for a Humanoid Robot"https://t.co/BnwJEW7DEv

0 0
12/27/2019
Placeholder
UltimApe@ultimape• over 5 years ago
Replying to @ultimape

For a school project I tried to build robotic legs out of Lego, so that one day I might let 'Cog' walk.https://t.co/beHrbax6hX

0 0
12/27/2019
Placeholder
UltimApe@ultimape• over 5 years ago
Replying to @ultimape

An idea that I've got stuck in my head: how one goes about teaching a robot that pulling on a string to tug a toy car vs pushing on a string. That they are two different things and One is not effective way to move said car.Or how to learn & model knowledge in a human like way.

1 0
12/27/2019
Placeholder
UltimApe@ultimape• over 5 years ago
Replying to @ultimape

"Statistical semantic models (SSMs; e.g., Landauer & Dumais, 1997) have recently been attacked as implausible cognitive models because they learn from only linguistic information and are not grounded in perception and action"https://t.co/5nADknsvhg

1 0
12/27/2019
Placeholder
UltimApe@ultimape• over 5 years ago
Replying to @ultimape

from google: https://t.co/vJIFySKDIPfrom facebook https://t.co/u5r6Xm88IN I find these funny: I've been thinking about this since before I was in college.It's why 'how collectives of ants/neruons model knowledge' is so intuitive for me: I've always been rooted in embodiment.

3 0
12/27/2019
Placeholder
UltimApe@ultimape• over 5 years ago
Replying to @ultimape

An exercise for the reader is to consider what dimension is missing when trying to 'translate' a static image into a series words using "visually grounded semantic representations".

3 0
12/27/2019