AI For Beginners

Today I begin my journey into learning the fundamentals of the global phenomena that is AI. Follow me as i explain today the importance of Data as it's used to train an AI system.

Now everyone at this point has some Idea of what AI is and what it can do. While some may have an inflated Idea of what AI can do, it is important to understand how an AI works in order to also understands what it can and cannot do. In this short blog I will explain a few simple concepts on how AI models are trained in order to better understand the limitations of AI.

The capabilities of an AI are determined completely by the sophistication and amount of data used to train it. AI systems learn by being given a large amount of data and enabling it to find patterns and make correlations to achieve a correct result. It is important that a diverse range and a plentiful amount of data is provided in order to prevent what is know as overfitting. 

Overfitting is a concept in AI programming that describes an issue in which the data provided to an AI is not  broad enough for an AI model to be smart enough. This can result in the AI making mistakes or having certain bias when faced with examples or problems it may not be familiar with. It is important to remember that an AI model can only be as smart as the data used to train them.

So in conclusion, AI is a tool that learns based on the boundaries of the data it has been provided with. It is good at finding correlations and patterns however it is unlikely to be capable or original or abstract thought, at least for now. It's important not to get lost the craze that we see online regarding AI and take to some time to see the actual uses and benefits it can provide to our society.