Machine Learning— what it is, why it matters, and how to get started

On a fateful day in April, I decided to plunge into Machine Learning (I didn’t know what the domain even entailed before that). Maybe I was bored. I was perhaps looking for something to do apart from App Development, which I have been doing since 2018. Four months and many all-nighters later, I feel like I have been enlightened! Today, I am a TensorFlow Certified Developer and a Google Cloud Certified Professional Machine Learning Engineer (a mouthful, I know), along with being a Google Certified Associate Android Developer (I have been one since 2018).
I have wanted to write about my experiences and journey for a while now. You see, ever since I received my TensorFlow Developer Certificate in July, my LinkedIn DMs have been flooding with all kinds of questions (from how I prepared for the exam to how to break into the field of AI) from a wide variety of people. As time went by, I began to notice some solid patterns and similarities in those questions. At some point, I realized that everyone had pretty much the same questions in mind.
I suffer from imposter syndrome, big time. I often questioned myself as to how I, a newbie, could write something of value online. I have seen many blogs written by very experienced and competent individuals, some with decades of experience. I have witnessed in-depth articles with code snippets written by the engineers that made the software themselves. And I wondered about how my journey could mean anything to anyone.
And then it struck me! I realized what needed to exist but didn’t already. And that is a human-centric approach to the questions that people have about this field. Most people only seem to talk about how great AI is, how AI is the future, and how much money you can make as an AI engineer. Most blogs provide you with tutorials on some framework or tool.
But I haven’t seen a single post about why my life will be better as a result of learning all of this and why the world even needs more people in this domain. We have a surplus of unemployed engineers right now, don’t we?
The first of this four-part series is to address all of these concerns. I don’t want to fill you with information that you can anyway get after a couple of Google searches. Experientially, I want to show you the journey I have been on and leave you to form your own conclusions.
It is Everywhere, And I Mean Everywhere
It may sound cliché, but I will say it anyway- AI is the future.
Sorry, I take that back.
AI is very much a part of our present-day reality. It’s like oxygen; you’re consuming it even when you aren’t aware of it. No need to panic, though; the apocalypse has not begun.
If you’re like most people, you own a smartphone. Your camera app may be using AI to take Portrait Mode photos. AI algorithms control the rate at which your battery is exhausted, which apps you see at the top of the drawer, and the best times to update your apps. The Google products you know and love- Search, Maps, GBoard, etc. are powered by AI. You’re already living in a science-fiction reality; you just haven’t realized it yet.
Our current education system is one where young people are pushed into the field of Computer Science, whether they enjoy it or not. As a result, there seems to be a general repulsion towards careers in tech. AI is different, though, because it is a tool, not a technology. Does your passion lie in singing? There’s a career in AI for you. What about dancing? Writing? Psychology? Finance? Trading? There’s a way to apply AI to it. What this means is that AI has blurred the line between non-tech careers and tech careers. You can be passionate about cooking and still be a part of the AI-community. This is fantastic news for two types of people- those who were forced into a tech career but had other interests, and those who work in non-tech domains and want to upgrade the way they work.
Another common complaint I have heard is from people who "don’t want to sit in a tiny cubicle in front of their computers and code all day." Well, I have good news here, too. AI is the kind of domain where you may have to write a total of 25–50 lines of code, which may require you to sit in front of a computer for about half an hour. Of course, there may be hours of thinking that go behind that half-hour, but you can use that time to smell the roses or do something else.
The last concern I’ve come across is whether ML is too mathematical. To that, I say, it’s up to you. If you’re working in a domain that requires you to use specific tools and frameworks, then you needn’t ever concern yourself with the intricacies of the mathematics behind them. If you want to be in a research-based space, you will need to work with Linear Algebra, Calculus, and Probability.
Still Not Convinced?
Perhaps you fall into one of the following two categories:
- You’re interested in tech, in general, but don’t know whether ML is the right space for you to be in. I suggest following my 1-month strategy- take up any domain in Computer Science (say Embedded Systems), make a list of all the resources you will have to go through before being proficient at that field, and then budget a month to do it. If you enjoy what you do, stick with it for another month. If you aren’t enjoying it, move on to another domain (say Internet of Things), and so on. Keep doing this until you find what you like. If ML is/isn’t your cup of tea, you will find out using this method. The budgeted 1-month is the commitment period, where you don’t give up, irrespective of how much you may dislike what you’re doing. If nothing else, you will have explored 12 domains within a year, and that’s excellent by itself.
- You’re not interested in tech but find yourself pushed into the domain. Get into AI! Find a way to make your field of interest (say, photography) about AI. Maybe create a tool that automatically edits a photo you take to look better? Your imagination and resolve are the only limitations here.
"My Laptop Is Garbage!"
This seems to be a massive concern for a lot of people who want to get into AI. From my experience, though, everyone who has approached me so far has had a better computer than me (mine is an ASUS X556UQK with a 7th gen Intel i7 processor, Nvidia 940MX GPU, 8 GB RAM, and 1 TB HDD).
Are you still feeling uncertain? Did you know that you can do AI programming on the cloud with Google Colab? Did you know that Google Cloud Platform and other competing platforms (Microsoft Azure and Amazon AWS) provide you with low-cost cloud resources that you can utilize on-demand? You have nothing to fear here.
And if you have the budget for it, you can buy a gaming laptop (those work just as well for developing AI applications). There’s also the Lambda Labs Tensorbook, a laptop built for Deep Learning Engineers (and my dream laptop).
You’re Already Obsolete
This is a problem you will face as an AI Engineer. The field is extremely new and moves very fast, and every single day, something is changing. What you’re an expert at today will become obsolete in half a year. Being successful in this field requires you to commit to being a life-long learner. Don’t look for a career in AI if you’re scared of change (More of this is covered in Part IV).
Is Having Too Much, Bad?
When I first embarked on my journey with AI, I had no idea where to start! There are many courses from Coursera, Udacity, Udemy, EDX, YouTube, Fast.ai, Nvidia, Intel, other online sites, online blogs, coaching centers, Universities, and in-house company training. I found the sheer number of resources available daunting, with immense pressure to pick the right ones (some of these options can be incredibly expensive). Many weeks of research led me to the path I finally took. Find out about my learning journey, the certifications I gained along the way, and my mistakes in the next parts of this series!
Part I: A Humble Beginning With AI
Part II: The TensorFlow Developer Certificate ▶
Part III: Cloud Professional ML Engineer ▶
Part IV: The Machine Learning Journey ▶
If you liked what you read, please share this piece. Subscribe to receive notifications about future articles. If you have questions, suggestions, or feedback, let’s discuss in the comments!
This article was originally published here.