Tutorial Hell: One surefire way to escape it!
Read Time ~ 7 Minutes
If you have ever tried to learn anything complex by yourself, you may have experienced “Tutorial Hell”
Tutorial hell is basically a term used when you can’t seem to break free of the following thought-action sequence:
I want to learn X.
I find a tutorial that covers X, go through it, but also includes Y so I guess I should learn Y too.
I want to learn Y to fully understand X.
I find a tutorial that covers Y, go through it, but also includes Z so I guess I should learn Z too.
Repeat ad nauseum.
This scenario has happened to me countless times in the past but by incorporating one specific technique, I have gotten better at not getting stuck in this rut.
What’s the one technique that I use to break, or altogether avoid, this vicious cycle?
SMART learning goals!
If you haven’t heard of SMART goals that’s ok. It’s basically a way to structure goals to give yourself the highest chance of achieving them by following a formula.
That formula is the acronym SMART and it stands for:
Specific
Measurable
Achievable
Relevant
Time Bound
So, let’s say you have the following learning goal. “I want to learn Deep Learning and finish a project for my portfolio”.
While this goal is something great to shoot for, it isn’t really SMART. So, let’s change that!
Specific
“I want to learn deep learning” is very vague. Deep learning is a HUGE field and covers a wide variety of topics. Narrowing this down will help us scope out what we really want to learn. If you are completely new to the topic that you want to learn and can’t be more specific due to lack of knowledge, that’s ok. Take some time to research the topic and try to figure out what exactly you want to learn. Just know, research is not the same as learning in this case. We want to figure out what to learn, not start!
An example of something more specific would be “I want to learn to implement a simple MLP deep learning model in Tensorflow that classifies images of handwritten digits using the MNIST dataset.”
Measurable
“finish a project for my portfolio.” is measurable but again, it’s a little bit ambiguous. What does finish mean? Is that a Jupyter Notebook with a bit of markdown or a slideshow that could be presented to a company? Do metrics matter? Are you trying to beat baseline or shooting for a specific number? Or maybe your code running without any errors is enough to be considered finished. You must define a finish line in order to cross it.
An example of something more measurable would be “a project for my portfolio that includes a Github repository, a README explaining the project, a Jupyter Notebook of the model, and a slideshow suitable to be presented to a board of directors.”
Achievable
This usually goes hand in hand when developing the Specific and Measurable portion of the goal. Two main things to focus on in this section is ability and scope. If you are just beginning to learn about a subject, your ability to learn/complete very technical and advanced materials related to that subject will be low. Your learning goal should reflect that ability and it’s ok if its low. We all have to start somewhere!
Scope is another aspect to consider. Saying you want to learn “all” of deep learning is not feasible. Understanding how much you can take on is important when creating a learning goal. It’s ok if you get the scope wrong at the start as you can always change it later!
Relevant
Ask yourself why you are trying to learn the thing you want learn in the first place. It should serve a purpose and there are many reasons for learning. Is it to help your career? Trying to shoot for a new job? Get a promotion at your current job? Just be more well rounded in your craft? Or are you learning a subject out of curiosity or general interest.
There isn’t really a wrong reason for learning, but you need to understand why you are doing it in the first place.
Time Bound
If procrastination affects you as much as it does me, this one is extremely important. Setting a deadline is crucial to be able to judge your progress on learning a new skill. It allows you to budget your time and keep the task realistic. This deadline is not set in stone, but it does give you an idea of how much time you need to devote to learning.
If you work a full-time job and have a family, maybe setting a really short deadline and then trying to spend 4 hours a day learning won’t work. Instead, allow yourself to create a manageable schedule for learning. This is a marathon, not a sprint!
Wrapping it up
Throughout this process I have tried to stress that your SMART learning goal is not some immutable thing. As you learn more and work on your goal things change. New information comes to light and you might have to change some aspect of your learning goal. Just try to ask yourself if the change is necessary in order to complete it!
Like I said before, I have been stuck in tutorial hell countless times and I found that incorporating the SMART framework really helped me work through it. To tell you the truth when I first learned about SMART I thought “How is this dumb little acronym going to help me?” but I gave it a shot and couldn’t be happier with the results!
I hope you enjoyed this edition of AI Insights.
Until next time.
Andrew-
Have something you want me to write about?
Head over to the contact page and drop me a message. I will be more than happy to read it over and see if I can provide any insights!
