Is coding required

As Nydia says, there is always some coding involved when you build solutions to ML problems, but you don’t need to code all the algorithms “from scratch”. The state of the art is so complex these days that even people who do this for a living typically use frameworks like TensorFlow or PyTorch and libraries like matplotlib and a hundred others. But even when you use TF, you’re calling TF functions, so that does involve coding in python. Programming is a skill that can be learned. There are many python courses on Coursera, including general ones like Python for Everybody from University of Michigan some that are specifically tailored for ML and Data Science practitioners, e.g. Python for Data Science, AI and Development from IBM.

Of course I hear that the LLMs can now program in python as well, so it’s entirely possible that this will get easier and easier as they continue to get better at that. Even if they eventually take over, it will still be a useful thing for you to develop your own skills in this area. You still need to know how to debug when things don’t work the first time and that’s a big part of what you learn when you become a programmer. Just writing the code is only the first step: you’re not done until it actually works correctly. It’s important not to treat debugging as an afterthought or an annoyance: it’s a key part of the job and worth the attention you will give it.

1 Like