Out of TV and remote which of them do you think is a main masterpiece of engineering?
You surely can change to thousands of channels and move things here to there using the remote of a TV, You are only limited to those 1000 things.
Many new ML engineers are nowadays stuck in this only
Give me a Biggest database you can find, I will ask chatGPT to give me code, then i will ask chatGPT to correct the code, then i will abuse chatGPT for wrong code and somehow at a point where everyone lose the hope the epoch started Walllahhh we created history
They were building image recognition 🙃
Google Tensorflow people are smart they have made the TV for you to play with and you are just using the remote only and feeling happy, wherein you should have been like this
Similar is with any framework and library they might have made it easier for you to develop an application, But with that ease comes a boundation to what extent you can go.
Indeed, the age-old mantra in the world of programming
You optimize for time complexity, and in return, you generously gift-wrap some space complexity.
When was the last time you saw someone making a react application or a tensorflow model and you were like WOOOOOOOOOOhh, Now that’s real shit.
Think think that might take you a while
Found any ?
Most of the things can be replicated easily because everyone can use those libraries and frameworks and might make an exactly same thing, Since most of the things are already available in the the form of packages it’s like
Plug and play
(Serious part) So where is the real software development in which you learned so many data structures, design pattern, DBMS and so many more trust me any new graduate will come to your face and will replace you with just your framework knowledge and that might hurt, Every developer wants to be the linchpin, but who worries about future layoffs when you’re firmly attached to these favorite frameworks?
Think it yourself, 10 crore + React developers how will you make your place in it and be unique because react is the TV and you have 1000 channels only to move around with 10 Crore+ people more, Now make something new and awesome.
Someone told me ML has got a lot of intensive maths, formulas, stats and equations, which even make it difficult to start but I have seen so many Jupiter notebook’s of students but the thing that was so common was that they hardly used any maths or just some basic maths in there notebook’s that too because chatGPT did it for them, Surely one of job chatGPT will take will be of those.
But it’s not even there problem tensor flow has already done those maths and equations for them, So the same problem arises over here is you never get to implement those maths and other knowledge which you learned. One question might burn many ML engineers
Do you use write ML algorithm or just use ML algorithm ?
No, that’s not what I meant.
If you are knowing Object oriented programming you might hear of the access modifiers like Public and private class, Online Tutorials and articles have created a new modifier known as
Internal implementation
In which what they really mean to say “Hey i don’t know how it’s working behind the seen and how it’s implemented, but i will say internal implementation because i am smart.”
Whose job is it to understand how the things are working behind the scene ?
Because behind Tensorflow or React or any other library lies so many layers of abstraction which makes it tough for anyone to understand that code, But trust me slowly and slowly as you keep diving in those layers of abstraction you will see the real software development, That’s where the real development is.
Which will actually make you a better software developer and not a frameworker,
In the vast sea of life, you find yourself floating on its surface, but your true purpose beckons from the unfathomable depths below.
So if your a frameworker like react or angular now don’t stop using it just start to get in implementation by reading the actual code and someday maybe just make a framework to yourself and
ML engineers just try creating few things from very scratch and not libraries where that real maths and fun hides, and i know someday you might even create a neural network of your own.
So don’t be a frameworker be what you truly are a Software developer 🫡
All the very best if this article forced you to think, I think i did my job well.
I hope you enjoyed it and learned a lot from this story, if you don’t wanna miss anytime I publish another story make sure to follow me and subscribe to get email updates.
You can follow me on Twitter and LinkedIn wherein I share much more information.
If you think you have a better or different perspective about this article don’t forget to comment I may include it in this with your name.
Be the first to comment