Whether you work as a software developer or not, you hear about artificial intelligence and machine learning almost everyday. Vehicles driving autonomously, governments tracking people or systems that recommend books and movies. It’s all around us.
All these things sound enormously complicated, creating an impression that only the best educated, brilliant and creative people can work on it. However, if machine learning is basically everywhere right now, there must be a lot of people capable of creating intelligent systems. Why it can’t be you?
No way, that’s too difficult for me
One of the most common excuses is that machine (or deep) learning is too difficult for mere mortals.
But hey, it’s not only Elon Musk creating autonomous vehicles and it’s not only Netflix that can recommend you a movie. Sure, their solutions might be inimitable. Well, you may not be able to create something better right off the bat. Nevertheless, it doesn’t mean that you’re too weak or unexperienced to do it. And I’ll tell you - you’re not.
I have no qualifications, I don’t like math
When you start reading about machine learning or even buy your first book, it often starts with a list of prerequisites - topics you should already understand before you continue reading. What you can see there is often math (a bit of statistics, linear algebra, calculus or maybe a probability theory).
At that point, plenty of people stop. And quit. Because they don’t like math. Or maybe it’s not that bad, but they are not good at it. Because they want to develop intelligent software, not to calculate integrals all over again.
They are wrong. You can explore machine learning tools and create your own software even if you’re not the best mathematician in the world. In fact, there are thousands of tools and high-level libraries that help you to develop your AI solutions - you won’t even notice they use complex math underneath.
In fact, you only need couple of lines to load data and train your first neural network, e.g. in scikit-learn:
iris = datasets.load_iris() X = iris.data y = iris.target model = LogisticRegression() model.fit(X, y) y_pred = classifier.predict(X) accuracy = accuracy_score(y, y_pred)
Well, that model is far away from even being a good one, but that’s it. Few lines of code to load data, train a model and even check how good it is. No matrices, integrals, statistics or whatnot. It’s fairly simple, isn’t it?
I don’t have time to learn Python and all these tools
Right, but is it really worth it?
It is definitely worth knowing a field that changes our life on daily basis. As a software developer, sooner or later you will also use ML in your projects. It would be good to have a good understanding of it by then.
Is it only a seasonal thing? Or will it stay with us forever? Nobody knows, but take a look at this graph from Google Trends. You can see a growing interest in topics like “machine learning” or “data science” over last decade (2010 - 2020).
Will this growth continue? As I said, nobody knows. I believe that there are still areas where we could introduce AI or where we could improve existing intelligent systems. In this article with statistics about AI and machine learning, we can read that people observe how lives and companies improve thanks to AI.
So I think this is not a seasonal thing and it’s totally worth it. And what’s your opinion?