How to prep for the upcoming AI Revolution
As usual, this post will have *actual* actionable information, and not be filled with a bunch of feel good fluff
This will be one of the most important posts for all Data professionals this whole year. This topic was requested by 2 paid readers:
There’s been a lot of talk about how AI will replace millions of jobs. And, how senior data professionals will no longer hire junior roles, but instead use AI. There’s also been plenty of “research” on what roles will be most likely to get rekd in this upcoming AI revolution. And, as we all know, research that is sponsored by an entity is never wrong:
Let’s talk about how you can start prepping for this AI revolution in the best way possible.
Table of Contents:
Roles & Responsibilities to Avoid
Using AI Tools Effectively (Actionable Advice)
AI Tools & The Impact on Interviews
Why Junior Devs Will Still Get Hired
Obviously, this post is *only* meant for those in the data industry.
1 - Roles & Responsibilities to avoid
1.1 Roles to avoid
Business Intelligence (BI) Analyst
In the past, when a company needed a guy to run some sort of forecasting/prediction, they would reach out to the BI Analyst. This guy would fetch some data, and then construct a complex model for it. Then, return his predictions to the executives.
Sound familiar? This is what a Machine Learning Engineer does today. The main difference is that the MLE provides the executives with an API/data pipeline. This API can be used to get predictions immediately on the get go.
The BI Analyst was already close to collapsing, with more automation coming, you need to avoid this role if you still want to be able to pay your bills.
I used to see job postings for Reporting Analyst quite a bit when I started university. It was actually difficult as hell to find 1 opening today. Anyways, here’s the requirements to be a Reporting Analyst:
In other words, this is a mediocre data analyst. I wonder why there’s so many openings for Data Analysts now, and barely any for a Reporting Analyst.
Guess it’s like “How many licks does it take to reach the center of a tootsie pop?”
Ans: The world may never know.
1.2 Responsibilities to avoid
If you taken any course on Machine Learning, then you know they’ll teach you the technical skills for about 10-20% of the time. The other vast majority is spent on things like:
What does this parameter mean
How does the objective function change
Using partial differential equations to solve equations
Unfortunately, in the real world… No one cares. Yep, this is why 99% of ALL Machine Learning courses = trash. 80-90% of what you learn is considered junk, and no one cares.
As of today, a machine can do a better job at building a Machine Learning model, than a human can. And, this trend will not reverse anytime soon.
If you think you’re going to be a guy, and let others fetch the data for you, and all you’ll have to do is build a model to collect your paycheck. This is the only kind of check you’ll be collecting:
Middle Manager - Idea Generation
There was a middle man role I saw once where the guy sets up meeting with sales teams, and reads research papers all day. Apparently, he was the “idea guy”. If you’ve worked with people before, and someone says they can’t do the *actual work*, but they can come up with great ideas. Avoid them like the plague
I’ll show you in the next section on how companies are tossing out most of their middle manager idea guys.
2 - Using AI Tools Effectively (Actionable Advice)
I’m not gonna waste your time talking about things like GPT4, co-pilot, etc.... Most of them give us generic answers, and if you presented them to the team leads. They’ll pat you on the back, and then say “lol no”
So, instead here are some examples on how AI is *actually* being used in the industry today. And, a list of tools to keep tabs on.
This section is only for the paid readers, the alpha presented here will *only stay here*.