1. Which two instructions in the "programming language" of the 2011 HW would be the most difficult for robots to follow?

Answer: 
		After step 8 , when the I affector is removed, It will be tough for robot to manage it's stability. 
Even if we assume that it is stable on that point, then we are again imposing I affector on a larger distance p coordinate ie. 5 cm
This balancing would not be very good because It was already in a less stable position difficult to hold it's axis intact. 
Torque could possibly change it's direction and axis.
		Also in step 7 , Rotation around the z axis by maintaining only two contact points and applying torque to rotate along the z axis
would be very tough. Normally when we rotate a pen in this way , we use a lot of contact points to balance the pen and maintaining axis of
rotation intact.
		Step No 13 is really tough step. It is really neccessarily to accurately estimate the force needed by the robot to maintain the 
balance with friction between the surface. Also when we write , we also rotate the hand position in the perticular direction which also 
helps moving in right directions. But here this won't happen because we are maintaing it in straight position (z axis). 
For forward movement, relative motion between I , T and M affectors is necessary which is not so accurate in this robot.

2. The robot following the learning paradigm as in Kalakrishnan is clearly gaining some expertise. Which aspects of the execution may be 
called implicit or automatic, and which aspects may be more explicit? What could be the "chunks" in this structure?

Answer : 
		Explicit part: Some of the motions are well defined and could be learnt explicitly eg. Finding coordinate of pencil, all the 
dimension measuring tasks , Movement in the vertical and horizontal direction etc.

		Implicit part: Robot needs implicit learning in maintaing adequate contact force and torque required. 
So a lot of training (implicit) is needed. It would also be needed in maintaining necessary amount of pressures and maintaing accurate
contact force between point of contacts. Slipping motion between M , I , T is also needed to learnt implicitly.
		For implementation of this part some machine learning algorithm has to be used. Above mentioned parameters will be learnt by this 
algorithm.
		
		Chunk: Here parts of instruction could be called a chunk as these are the parts that will be frequently used by robot , making them
easy to be recollected. eg. Perticualr moves could be called chunks.

3. Comment on whether human learning may also be following similar "reward" based processes? Consider the learning process for the
fire-fighting expert who knows how to fight complex fires. 
Answer:
		According to my view, Human learning might follow similar "reward" based process. We had a really long debate discussing what
could be the possible reward in the tasks which humans do noramally. Some physical processes of human body might not be functioning on
this way. eg pumping of heart and flow of body in humans. But as far learning is considered, satisfaction and internal motivation is
 really important. 
		Constant positive feedback is also important in learning.
		We also discussed an example of a child. Sometimes when a child is doing some bad thing, we usually scold them and say him 
not to do again. This has a impact on him, This negative feedback of that perticular task enforces him to not do that thing again.
		Similar with positive feedback.
		
		In starting , a fire-fighter just uses the knowledge gained through his training (Explicit Knowledge) , Then depending on the
positive or negative feedback , he changes his abstractions of previously attained knowledge. Internal satisfication also 
inspires him to carry out his task with greater accuracy. On constant negative feedback a fire-fighter might leave his job, 
Though he might be enjoying the task earilier but later negative feedback will change his mind set completely.