Artificial Intelligence has shown excellent results in mimicking some aspects of human intelligence, like object detection, game playing, and text generation. However, incorporating the ability of backing actions with reasoning, similar to human thought process, remains a challenge. A novel project by scientists at the University of British Columbia and Vector Institute introduces an innovative technique called 'Thought Cloning', aiming to equip AI systems with the capability of understanding and implementing human thought process parallel to actions. The concept of Thought Cloning essentially incorporates the reasoning behind certain actions during the AI's training period, so that the AI model can generate a similar reasoning process for its own actions. This differs from the conventional 'Behavior Cloning' that merely imitates human behavior, failing to explain the reasoning behind those actions. Thought Cloning promises a myriad of benefits - efficient training, easy troubleshooting and fixing, and prevention of harmful actions. The process of Thought Cloning leads the AI system to predict the correct reasoning and actions to take in real-time scenarios, allowing it to be more flexible and reliable. Although presenting promising potential, introducing Thought Cloning in real-world AI brings about its own set of challenges. The core being creating the training data - generating explicit narration or reasoning for every human action and behavior. The authors suggest using YouTube videos where people explain their tasks, but even then, implicit human behavior that can’t be easily explained remains a hurdle. Despite the complexities, Thought Cloning represents an exciting direction in AI research, capable of building a robust connection between AI actions and reasoning.