Ask HN: Is data science a dying profession? | Hacker News Reader

Ask HN: Is data science a dying profession?

18 pointsusgroup12 comments

In my opinion, the term "Data Science" is becoming vague. Similar to the term "Web Developer" which lost its clarity in the 2000s and was divided into various roles like backend, frontend, UI, UX, app developers, and so on. Currently, "Data Science" has expanded and there are multiple subfields that have developed into their professions like data analyst, data engineer, machine learning engineer, machine learning scientist, and more. So in that sense yes, it's dying. But dying of success.

My understanding is that data science implies big data.

As Jordan Tigani observes, there are not that many really big data situations[1].

Thus data science proficiency might be in support of business initiatives in a particular application domain and not a means to an end in itself.

[1]:https://motherduck.com/blog/big-data-is-dead/

> My understanding is that data science implies big data.

It really doesn't.

I dont think OP is wrong -- the role popped up after 2007 Hadoop release. It was almost synonymous with running ML on big data and data scientists were typically computer science PhDs. It has become something quite different though. Now everyone and their dog is a data scientist.

Big data and data science became popular terms around the same time, doesn't mean that one implies the other like the GP comment suggested. Big data isn't a concern for most working data scientists today - not all datasets are that big, even cheap machines are very powerful and in the cases when you are working at huge scale a lot of the complexity is abstracted away from the data science work anyway.

so typically there are two sides of the house

1. data engineering - dealing with etl pipelines, data warehouses, data lakes, databases

2. data science - ai/ml, models etc

#1 preps things so #2 can do their jobs

Not at all, why would it be?

Increasingly a subskill of other positions, automated, or specialized.

The specialized part is kind of true to some extent but that's getting bogged down into job titles - data analyst, applied scientist, ML engineer, research scientist - some of these are used more often because "data scientist" is too broad and can mean a bunch of different things. But titles, schmitles - the point is that the skills are still very relevant.

And I'm not seeing anything getting automated.

Agreed. The title data scientist is probably going to go away and be replaced with more specific titles you've listed above.

IMO, the emergence of specializations is a good thing because a) it means the data field (and the tools it uses) are maturing and b) you can have a more focused career trajectory instead of feeling the pressure to learn EVERYTHING

I think this is correct. So a DS does a bunch of stuff.

- Some of the stuff becomes diffused and common. Anyone can do it.

- Some of the stuff becomes data engineering.

- Some of the stuff becomes fitting a model -- anyone can do it.

- Some of the stuff ends up subsumed in a cloud providers offering and mostly automated.

- Some of the stuff becomes a SaaS (e.g. A/B testing).

Death by factorisation.

Specialisation is another sort of death, where X doesn't need a "Data scientist" any more, they need a "X data scientist". A bit like how "scientific computing" is just programming, except you wouldn't hire "just" programmers to do it -- you hire programmers who specialise in scientific computing.

I think is growing more than dying!

Reply on news.ycombinator.com