I am going to assume you want to learn the “Data Science” side of things.

This includes Google Deepmind, IBM Watson, and Amazon Alexa’s work.

Then, if you are a Non-IT, I would say — maybe.

AI has a lot of maths involved. If you know uni maths including some linear algebra (inverse, determinant, trace, differentiation of matrices, etc) + statistics (MLE, Confidence Intervals, Linear Models) + Calculus (Several Variable, Chain Rule, Newtons Method), then you should be fine.

Programming is the core of AI work. You will need to learn Python, which is “easier” to understand than most other languages. HOWEVER, this DOES NOT mean Python is “easy”. Python has lots of tricks and tips involved, and it gets exceptionally slow when doing DS and AI work. That is why you need to know maths to vectorize your operations and use matrices to your advantage.

In my opinion, if you have 0 programming background, you will find it very confusing at first to learn logical programming steps. Even worse, if you don’t have enough mathematical skills, it will make it even harder.

At the same time, AI work requires some of of “intuition”, a “sense of curiosity” and also communication skills.

Even though AI might seem all the “hype” and all, it is not easy. In fact, I would say it is like doing a uni degree triple majoring in Computer Science, Statistics and some Commerce degree.