DEV Community

Cover image for Dirty Jobs for Data Science
Max Ong Zong Bao
Max Ong Zong Bao

Posted on • Updated on • Originally published at

Dirty Jobs for Data Science


Go to places where there is little to no competition. Do not compete in a red ocean instead create your own blue ocean. Wise words from a book called Blue Ocean Strategy which I had read a decades ago.

In fact, I had been moving towards, creating and building these blue oceans having read that book. Recently, I had stumbled across an article in the local news. Which talks about Trade Schools/Vocational Institutions teaching their students AI curriculums.

I had encountered alot of positive comments and concerns regarding the relevance or ambition of teaching vocational/trade school students with knowledge of AI when I was sharing to various developer or data science group that I am in.

Why Learning AI in Trade School Makes Sense

First of all, to get the record straight. I do not expect these students to work as a Data Scientist. If that is your reason for rejecting on what they are doing, I am gonna burst your bubble on it. As being a data scientist requires a massive amount of knowledge, training, domain expertise & experience to do data science work.

Therefore I am not surprised that the requirements of a Data Scientist is to hold a Masters or PhD in their related field with an MBA. Unless they have proven that they are capable of working with a data science team. To build data products & the infrastructure to deliver value for the organisation or company.

So what makes sense to me is just exposure and the low hanging fruits that blend their skillsets for data science work. In areas that is not sexy for a data scientist or attract anyone to work in that data science work.

Doing work like data collection, data scraping, data wrangling, data labelling, data classification, building the data lake, ETL (Extract, Transform, Load).

These are essential tasks related to data science but delivers value. Despite you do not require much knowledge of data science to make it work. Often they are not sexy or glamorous. This is why I consider these as dirty jobs and not the blue-collar equivalent of it.

Low-Value Jobs

What if I were to tell you that data labelling/data annotation is worth a billion dollars. Will you be interested in tackling the industry? If not why? Is it not sexy, glamorous or prestigious to your taste?

What I expect from the students who are taught AI. They might be approaching it as a supporting role. Take for an example like a Data engineer, API developer, chatbot developer, IoT field engineer, Drone operator, data labeller.

With diverse skill sets that complement what is taught by the trade school or vocational institution. Due to their philosophy has been central to train graduates with relevant skillsets for the job that requires their skills. Much like coding boot camps focus on teaching and educating the students to become developers. The difference is just the value chain of how these graduates deliver value & their payscale in the value they provide to the marketplace.

Is It Ambitious to Teach AI for Trade School?

Yes, to teach the whole range of AI/Data Science curriculums of the level for PhD/Master is ambitious and crazy. But it does not mean that these students can not learn to contribute to the development & support of these AI systems.

Besides the last thing, I would want the students of these vocational institution/trade school is to be treated like a Mechanical Turk to work on menial or mind-numbing work. Instead, they might be interested in digging deeper for opportunities to develop their skills, knowledge and experience as a specialist for specific data science work.


I have covered why I feel about vocational institutions/trade schools teaching AI curriculums. I think of it as a complement to what they are learning now in the school to create, support and maintain an AI system.

I believe these graduates could be involved through low-value work that is often neglected, unglamorous or not sexy.

Take for an example data collection and data wrangling of raw data. Instead of just getting it from a source. That has done alot of hard work on data wrangling & labelling to make it usable for data exploration.

In terms of ambition for these trades schools/vocational institutions to train their students on AI curriculums. I think it will be more towards using the knowledge and complement with what they had learnt to use it for their work.

Compared to some coding bootcamps I know. Their aim is to train & educate the data scientist/data analyst. Those are the kind that does raise concern for me. Than any other trade school/vocational institution, who aims to prepare their students for the future where AI will be part of their daily lives. It is like how we are taking advantage of Google's search engine to help us to learn or solve problems with it.

Lastly, are you looking to specialise as a developer? If yes, I’m giving away my free ebook called "Picking Your Specialisation as a Developer". It is for anyone interested in commanding a higher salary or simply doing the work they like.

This post includes affiliate links, I may receive compensation if you purchase products or services from the different links provided in this article.


Top comments (0)