AI is already doing work at the junior programmer level; some predictions are that in 2-3 years, it will work at the senior programmer level, i.e., probably better and faster. On the other hand, it is surprising how many people, including programmers, still know little about this topic, which could soon bring big changes in the field they are working in, and in others as well.

People are mainly divided into two groups: those who have heard something, use ChatGPT and maybe some other tools, and that’s all, and the rest who pretend to know so as not to look stupid.

When you enter the field of courses, there is real noise and chaos; it is difficult to understand where to start quickly. Many people immediately gravitate towards “flashy” things like GenAI and various boot camps for AI agents because that seems the most practical.

I think one should first complete courses that provide a good theoretical basis for what is happening “under the hood.” There are claims that this is not necessary, but actually, it is because, without it, one is just a technician who does most things semi-automatically and is not an expert in this topic.

The recommendation for entering the Machine Learning topic is a course that does not need a recommendation: Machine Learning Specialization by Andrew Ng (Coursera) (I am not an affiliate, just so you know). I would single out this course from all the paid and free ones I’ve had experience with so far and put it in first place.

I enrolled in this course some time ago, because it was recommended by almost all authorities in the field, and two chatbots also recommended it to me. It had a surprisingly high rating (4.9). After that, it became clear to me.
Andrew Ng is one of the greatest authorities in the field. He explains complicated mathematical models such as linear regression, gradient descent, and cost function in a very nice, understandable, and pleasant way. If I had a math teacher like this in school, I would be the best in that subject. The lectures are facilitated by the use of a Jupiter Notebook, which contains short code snippets written in Python, using libraries like Numpy and Matplotlib, that the student activates and gets results and diagrams.

A very useful course, even an example of what courses on this topic should look like. Covers Python, Scikit-learn, and neural networks with hands-on projects. Ng’s name is a trusted credential. Ideal for entry-level roles.