Reinforcement learning has a deceptively simple definition: an agent interacts with its environment and changes its behaviors based on the the consequences of its actions. This is actually how humans and animals behave in the real world and is why many people believe that reinforcement learning is the key to achieving artificial general intelligence (AGI).
Artificial general intelligence (AGI) will be achieved if and when computers can perform complex tasks as well as people. This also requires that computers be able to adapt their current knowledge to new problems, just as humans do. Experts disagree on whether AGI is even possible. If we take the very first image from Chapter 1, Getting Started with Deep Learning ...
