joswilliams \ stacker news

@joswilliams

255 sats stacked

stacking since: #255477longest cowboy streak: 1

Data Scientist.

10 sats \ 2 replies \ @DarthCoin 16 Sep 2023

What exactly is a "data scientist"? Data = something very specific, that exist, is already proven, mostly by math Science = something to be discovered, something that can be debated, debunked, peer review. Example: I am studying a flower, I wrote a paper about it. Then you came and try show that my study is not complete, false or you just want to improve it with more info. All that "debate" is the science, something like a common ground where 2 people can agree. But that doesn't mean is really like they say, because can come up a 3rd one saying that their whole "study" is false, coming with new discoveries. Science as per is in a continuous change. Not like math (data).

So for me something doesn't add up in this. If you work with data, I think the more appropriate term could be mathematician, data collector or something like that.

271 sats \ 1 reply \ @sorukumar 19 Sep 2023

Broadly, I agree with the intuition behind Darth's response. It depends on which part of the elephant we are looking at.

My response comes in two parts. The first part dives into how those in the industry explain why the 'data scientist' role is named as such. The second part delves into the backstory of the birth of the role labeled 'data scientist.'

Part 1

Let's break down the term 'Data Scientist:

Data:

Data is, for sure, something specific. In the real world, events are happening, and to make sense of or to learn from those events, we want to capture those events. You break the events into parts and you capture those parts, or in some cases subparts, in data.

Person A and Person B, might look at the same event but define 'data' differently to capture its essence. There's an art to translating real-world activities into data or metrics. So, while data involves math, fortunately, or unfortunately, there is an art to math.

The map is not the territory. Likewise, the data you capture serves as a proxy for real-world activities.

Science: I'll go with what you have said.

Now, consider a company with humongous data. Two data enthusiasts in different teams gaze upon the same data.

Guy 1 says: This is what is happening with the product (Discovery) This is what we can do

The senior data guy/ management might nod in agreement or raise an eyebrow, asking for further exploration. The loop continues until everyone's comfy.

Guy 2 says: this is what is happening with the product This is what we can do

The senior data guy/management may give the nod or shake their heads, prompting deeper dives into the data.

The two teams meet, and The loop of returning to the data and crafting stories goes on.

The key point here is that this process isn't all that different from how we approach 'science.' People come up with different theories to explain the events of nature, and there is a debate.

The real story:

The 'data scientist' title was first minted ( or popularized) by LinkedIn around 2009. It was a combination of: 'the need of the company', 'the specific skills they sought for the role', and a bit of branding to attract the right talent.'

LinkedIn (DJ Patil who gets the credit for the term) was on the hunt for stats/maths grad or PHDs. Many of them were getting jobs in pharma often with the title of a scientist. Some of them were not super excited with the Humber data analyst label even with a heftier paycheck. So they got the title to do the task LinkedIn wanted them to do.