1. At the MIT robotics lab humanoid qualities can include fallibility. What is fallibility?

Question

Why might humans value fallibility in Robots?

Answer 1

To err is human.  Fallibility is the ability to make mistakes.  Check out a dictionary.  As far as having robots be more human like, I really don't know.  It's certainly not like there is much risk of getting lonely because the earth is running out of humans.

Answer 2

Ans example as to why fallibility would be of value:

Suppose that in the not to distant future it were possible to create robotic baseball players that were perfect.   Suppose that you played a game of baseball entirely with these perfect robotic players.  Once the novelty wore off you would realize that the resulting game would be boring.     Perfect pitches always being hit by a perfect batter resulting in the same marginal hit that is perfectly caught.   The results of each batter would be the same and the game would soon become very predictable and boring.

Now add fallibility into the mix the ability to not be so perfect. Now slight imperfections in pitching and hitting and catching combine to add a variety of results to the game.  By removing perfection the resulting game would be come more interesting and less predictable.

Another example would be a robitic musician.   Instead of playing perfectly timed notes the variations introduced by fallibility would make the produced music sound more natural.