YouTube Deep Summary

[00:00] Many of you want to be machine learning [00:02] engineers. I get it. It's a great job with interesting work, high pay, and overall, it's just quite cool. However, it's definitely not a walk in the park. And in this video, I want to give my candid and to be honest, brutally honest advice for those of you who are aspiring machine learning engineers. Let's get [00:19] into it. If you want to become a machine learning engineer, then you need to spend at least 10 hours per week studying outside your normal responsibilities. I am sorry if that upsets you, but if you want to land a job in probably the highest paying tech profession, you need to put in the extra effort that other people aren't willing to do. There's literally no way around it. Without sounding arrogant, I managed to learn something new in machine learning every single week. And that's with a full-time job, [00:48] uploading YouTube videos, writing blog [00:51] posts, and also exercising nearly every [00:53] single day. So, if I can find time to do it, then so can you. You just got to stop making the excuses and make it a priority. Almost everything I've achieved in my career came from consistently learning and documenting everything I'm studying along the way. I've written over 150 technical articles on Medium. These span topics like be [01:11] statistics, neural networks, mark of [01:14] change, optimization, time series. Literally such a wide range spectrum of topics I've covered. And through documenting and learning, it's really solidified my understanding of pretty much the whole field, at least the fundamentals. And that has allowed me to really excel because I'm learning things quicker than other people. Again, this is not for me to boast, but just to show you the level of commitment you need if you want to land a job as a machine learning engineer. Think of this [01:40] profession the same as lawyers, doctors, [01:43] accountants. These people study for years and practice for years to become qualified in that area. You've got to think of machine learning the same way. Just because it's newer, it hasn't got the same kind of, you know, you know, people think like, oh, it's so easy. Do an online course or qualification and you can become one. You can't. It's a [02:01] profession and it's a discipline. So you have to dedicate studying time over years if you really want to become a top level machine learning engineer. I often say that everything is actually easy but just hard. And what I mean by that is that it's easy to understand exactly what you need to do but it's just difficult to implement it over a consistent period of time and have that discipline just for like I said years on. Most people just don't have that and that's just being honest. So pick [02:28] something you want to learn in machine [02:29] learning. Learn it and stick to it. And after you're done learning it, recycle that process and learn something new. Again, there's no secret to it. It's just starting something, finishing it, and doing that again and again and again for years. If you want to become a top [02:42] level machine learning engineer, like I [02:44] said, if you're looking for new resource [02:46] to upskill your machine learning skills, [02:48] then I recommend Skillup by simply learn [02:50] who are kindly sponsoring this video. Skillup is a free resource offered by simply learn giving you self-paced courses on a variety of topics like AI, cyber security, software engineering, ASIS, machine learning, literally pretty much anything in the tech space. Many of these courses are developed in collaboration of top companies like Google, Amazon, and Microsoft. So, you'll literally be learning from the best. It is ideal for literally anyone. [03:15] Whether you're a complete beginner or a [03:17] seasoned veteran in the field, there'll [03:19] be something there that is viable to you [03:20] to learn. and you will even earn a certificate after passing the course. If you're unsure about what to choose, Skillup offers a tailored recommendation questionnaire that you fill out. It literally takes a few minutes and from this questionnaire, you recommended courses that suit your needs or kind of what you're looking to learn. So, there's literally no excuse because you can literally find something that you are looking for without realizing. Shown [03:42] on screen here is me walking through [03:44] this process and it literally takes less [03:46] than 2 minutes to do. I will leave all of this linked in the description below and I encourage you to browse their course catalog to find something that interests you and just start learning. Even with the most ideal background, it'll take you at least 2 years if you want to land a job as a machine learning engineer at a top company. Fall don't fall into the trap that taking a few online courses, doing a few simple projects will land you a job in this field. It really, really won't. Online [04:13] certifications and courses are really [04:15] good to teach you the initial [04:17] information and just basically expand [04:19] your knowledge of the field. To be a machine learning engineer, you need to know loads of things. Maths, stats, software engineering, machine learning, DevOps, cloud systems, all of these are literally individual professions in their own. So, you can't expect to learn all of this simply from online courses. It's just not going to happen. Most of [04:36] these skills can only really be [04:38] developed through real world experience. That's why I recommend people become data scientists or software engineers first for a few years to develop that skill set and then learn their other skills on the side and eventually transition into becoming a machine learning engineer. And even then to most people who want to be a data scientist I recommend being a data analyst first then data scientist then machine learning engineer. As you can see it's not a short or quick journey. It'll take like I said at least 2 years realistically three or four like it took me. So stop thinking that you can just [05:08] become a machine learning engineer right [05:10] off the bat. It's very rare. It might happen. But those people are probably not watching this video because they probably already know what to do. They probably come from top PhDs and top universities with the top level machine learning research. But that's a whole [05:22] different kettle of fish. I'm speaking to the most ordinary person who probably has a degree, STEM degree, works in, you know, looking to break into tech. You're not going to get there within a year. It'll take you 3, four, 5 years. Like I said, in my opinion, accepting the fact that it will take you years to become a machine learning engineer should be quite liberating because it kind of takes the pressure off. So just take [05:42] your time and learn things deeply into a [05:44] good level. Really understand how machine learning algorithms work. Understand how to write really good code. These things take time. They are skill. People dedicate their whole lives [05:54] to these areas. So like I said, take your time, learn things thoroughly, and eventually when the time is right, you will become a machine learning engineer. But you have to go through the process and go through all the steps. News flash, a machine learning engineer is not an AI engineer. So stop thinking that after you've built a basic chatbt API call that you're suddenly machine learning engineer. You're not because [06:19] that's not what machine learning [06:21] engineer is. And to be honest, that's actually quite easy to do and most people probably can do it without a tech background. As a machine learning engineer, you're expected to have a really deep knowledge of how the algorithms work under the hood. You need to understand maths, linear algebra, calculus. You also need to have a really good grasp of statistical learning theory which the backbone of all this this whole thing. Like these are really [06:43] technical deep skills that takes kind of [06:46] like like I said a long time of [06:47] studying. So simply calling an API to a chatbot, it's not machine learning engineering by any means. You need to know the core algorithms like linear regression, generalized linear models, decision trees, neural networks, and support vector machines inside out. Most people claim to know these very well, but you'll actually be surprised of how well you actually don't know them. I've mocked interviewed so many candidates, and you'll be surprised by a number who literally can't explain gradient descent from the first principles of calculus. [07:17] Again, I'm not being harsh or mean here. I'm just explaining the way I've seen things. And I think a lot of people think they know more than they actually do. I always tell people to stop rushing to learn all the flashy things like Gen AI, computer vision, NLP, reinforcement learning. You should actually spend your first few years mastering the fundamentals to a really, really high level because to be honest, most of the job is the fundamentals. Most of the job [07:41] is applying regression models, decision [07:43] trees, and maybe neural networks. But even then, that's a stretch. So understanding the basics of supervised learning, the theory behind it, how statistics works, how data works, these are the things you should spend your initial time learning because these fundamentals are literally the 80/20 of data science and machine learning. So spend your initial years learning the fundamentals to a really really high level that you're going to answer really deep questions about them. that is such a better investment of your time than learning Gen AI which to be honest in my opinion is a bit of a fad and and be kind of over in a few years. So things [08:17] that have stood the test of time like [08:19] statistics and statistical learning [08:21] theory are well worth investing your [08:23] knowledge in. If you want to test your fundamental machine learning and statistics knowledge then I offer mock interviews where I basically give you questions and scenarios that I faced in literally so many machine learning interviews at top companies. So, if that interests you, then you can book a call with me in the description below if you want to test your machine learning skills. Let's end with something that is just pretty obvious. Becoming a machine learning engineer is just difficult. [08:51] I've said it throughout this video. To become a machine learning engineer, you need to know so many disciplines. Not to mention, you need to dedicate a lot of time outside of work to upskill yourself. And this process takes years. So it is very hard to stick at something for a continuous period of time and put in the extra effort with to be honest not necessarily a guarantee at the end like you obviously will learn stuff but there's obviously no guarantees in life that you actually will land that machine learning engineer role and even with the most perfect ideal background like a STEM subject mast's degree or even a PhD it'll still take you a bit of time so even if you don't have that right background it's even harder it's not impossible but it is just a lot harder And that's just me being honest. I also [09:35] have a saying where I think anyone can [09:37] become a machine learning engineer, but [09:39] that doesn't mean everyone should become [09:41] an engineer or try to become one because [09:44] like I said, it's a long process, 2 to 3 [09:46] years at least. And most people are just not willing to commit to that time frame, which I totally get because there's also other careers that may suit you more. And there's alo other things in life that you want to focus on apart from grinding in the evenings and the weekends learning more knowledge, more machine learning to get that role. I totally get it. For some people, it's just not worth it. You just have to be [10:07] honest with yourself and think is [10:10] investing two to three years of my life [10:13] to becoming a machine learning engineer [10:14] worth it. And in fact, most of the time is more like four to five. And that's a very very long time. For me personally, investing three to four years to learn and upskill myself for a career that I'm going to have for decades that I'm going to really enjoy and really take pride in is worth it. That's me personally. You [10:33] just have to make sure or decide if that [10:35] sacrifice is worth it for you. Without going too much into the grind mentality, I'd like to leave you with a tweet from Alatmozi that I really like. Nothing is that hard if you try hard, which is why most things are hard for most people. They try so little. The harder you try, the easier it gets. So to me, that [10:54] perfectly sums it up. Now, if after watching this video, you're willing to take that extra time to become a machine learning engineer, then I recommend you check out this video where I detailed the exact road map I would follow if I was looking to become a machine learning engineer again. See you in the next video.