Update README

README.md

@@ -2,6 +2,7 @@
 This is as an exercise to practice with RL. I wanted to build the agent and the environment without any libraries.
 
 I choose golang because I am learning the language.
+
 I choose tic-tac-toe because in my foolish youth I coded an "AI" full of if that play tic-tac-toe. I wanted see how it would work with RL.
 
 ## Structure
@@ -14,7 +15,7 @@ Two agents play against each other. For the first part of the program, both of t
 | **o** wins **47.3%** times | **o** wins **99.2%** times |
 
 ## Observation
-It does not take long for an agent to learn. I first, after the second agent forgot what it has learned, I was letting it learning again. The consequence was that the distinction between the two agents wasn't clear.
+It does not take long for an agent to learn. I first, after the second agent forgot what it has learned, I was letting it learning again. The consequence was that the distinction between the two agents wins wasn't clear. That is why I also make the second agent to stop learning.
 
 ## TODO
-Make the output better. A graph showing the learning curves of the agents ?
+- [x] Make the output better. A graph showing the learning curves of the agents ?