|
Hey Reader, This feels weird. This is the first email I've sent since I paused this newsletter back in May. A few things have happened since then:
Recently, I have been doing several calls with a couple of you to learn more about the issues you are facing in data science and machine learning. I am currently building something that will help you accelerate your progress toward landing a job (stay tuned 👀). There is one common theme I see. WhAt cOuRsEs Do I tAkE nExT? Honestly, I love you all, but this question needs to leave the group chat. Once you have done one course on a particular subject, you're finished (finito, fini, fertig, acabado) whatever language you speak. You don't need to do three courses on Python to learn Python. You do one, then move on to practising and building. The same applies to all other subject areas: SQL, machine learning, statistics and deep learning. I get it. Taking more courses feels "productive", but in reality, it slowly becomes "unproductive" as it stops being the most valuable thing for you to do at your current position. At that point, you are in "tutorial hell." Loads of people are aware of this concept, yet many still fall into it. Sure, there can be exceptions where you can take more than one course and tutorials on more sophisticated topics. Deep learning would be one example where you could take separate courses on computer vision and natural language processing. However, they are technically two distinct subjects, so the point becomes moot. I remember when I was studying to become a data scientist, I followed this principle without even realising.
That's what actually moves the needle. So, if you are thinking of taking "another course", take five minutes and think to yourself. "Is this me procrastinating from doing the hard work of actually building projects and implementing my skills?" You'll know the answer deep down. Speak soon, ​ PS: If you're struggling with your learning roadmap or anything else in your job search, feel free to book a call with me here. PPS: For those of you outside the mid-20s London bubble, according to Google: "Splitting the G" is a Guinness drinking challenge where you take one initial, large sip of a freshly poured pint to align the dividing line between the dark stout and the creamy foam with the middle of the "G" on the logo. The goal is to make the line of the beer pass through the center of the "G," which requires a steady hand and a bit of practice" However, if you are like me, you don't need practise. Just pure inherent skill. |
Weekly emails helping you land a job in data science or machine learning
Hey Reader, Let's talk projects. These beauties are essential for landing interviews if you lack real-world experience. But, I am begging you, stop asking me for exact project ideas. This is not because I don't want to help you, but any project I suggest to you will immediately be a "bad" project. Let me explain. I have interviewed dozens of candidates at this point, and the projects that stand out are the deeply personal ones. The reason this works is that there is some "romanticism" to...