I love heavy metal.
I spent most of my teenage years in my room, practicing on my guitar Metallica’s songs. I had to play as fast and hard as them.
When I was 17, my family moved to another town and I had to change guitar teacher. By that time, I was able to play every Metallica song I wanted. I thought that it was enough to make me a great guitar player, but my new teacher Claudio didn’t agree with me.
As it turns out, most metalheads have no knowledge of harmony, and I wasn’t an exception. I had to fill this gap, so Claudio started to teach me jazz (a genre that requires everything I was bad at).
Playing jazz was a humbling experience. Most of the hard-gained skills that made me a good metal player were completely useless. I found myself fighting with complex chords and odd melodies, and felt like a beginner again.
I didn’t even like jazz at the time, and even after months of work, I was still a crappy jazz player. I decided to stop my jazz venture, and went back to playing my favorite music.
What happened later was surprising. By getting enough exposure to the world of jazz, I was able to add new elements to my usual guitar playing. Even if I wasn’t ever going to be a jazz player, the exposure I had to the world of jazz was an asset for me. Complex chords, harmony, dissonances, all new tricks up my sleeve. Tricks that few other metal players had.
I get daily messages on Instagram or LinkedIn from people that want to start a career in Data Science. People coming from all sort of backgrounds: business, marketing, law, engineering. The script is always the same:
“Hey Gianluca, I am a marketeer/engineer/manager/etc. I got passionate about Data Science and AI. I know nothing about coding and math, how can I become a Data Scientist?
I have absolute respect for people that want to change their career, and how couldn’t I? These people are telling me they want to:
- Learn to code from scratch. Doable, but hard, long, and frustrating (especially without guidance).
- Learn math and statistics. Also doable, but even harder than coding.
- Gain experience, starting from junior or even intern positions.
- Start spending most days in a room, alone, cleaning data. That’s what most Data Scientists do, and if you need to have the right personality for it.
- Throw in the bin most of your previous knowledge and experience in other fields.
- If you really want all of the above, go ahead, you have my blessing. The reality is that most people don’t actually want it.
Most of them are motivated by a genuine interest in the topic, a hungry job market and good pay checks. All fair reasons to me. Yet, I' don’t think that becoming a Data Scientist is the way to meet their goals.
Companies are not interested in building amazing tech. They want to use good technology to solve their challenges. In this context, Data Scientists are just one of the figures needed, and arguably not even the most important. What they need, is someone capable of identifying opportunities, communicate with technical teams, and run the projects.
Do we care if these people know how to code a Neural Network? Definitely not. Such a key player in a company’s AI journey has a different set of skills. This person needs to have an operational understanding of AI, together with deep knowledge of his specific industry. In this context, having a marketing/business/law/whatever background becomes an asset. Not knowing math or code stops being a limitation. Endless opportunities arise.
I’ll let jazz players play jazz. I’m a metalhead and I’ll keep playing heavy metal, using all these jazzy chords I know to be better than my peers.
In the same way, I hope you’ll let Data Scientists do Data Science. Focus on your strengths, and learn enough of AI to become the new figure every company needs.