Why does data science get paid less than software engineers. Ai/ml get paid the most and data science is related. #data #dataanalytics #datascience
Supply and demand. If you open any DS position you are guaranteed to get hundreds of qualified candidates in a day. The same isn't true for most SE positions, in particular more senior roles.
i thought SE roles are super saturated with candidates though
They are saturated, although only few % are actually qualified for each opening. But nowhere as saturated as DS, as pretty much everyone signs up for DS. Not just CS or Stat majors, but also psych, sociology, finance, math, business, natural sciences, engineers, and on all apply to it. DS is a job anyone can do but very little people can do well. The process around selection is poor in filtering experts from quacks, resulting in many false positives and the massive dip in DS hiring right now and extreme oversaturation of applicants.
You know how many people want to just build models without ever running unit tests or writing config files? A lot!
Supply and demand. Too many people flow into DS
I keep hearing this but 90% of the applicants are horrible. It’s not that there’s too much supply it’s that people thing data is easy (easier than SWE) but in reality they’re nowhere close to qualified
Data science core knowledge + systems programming ( bash, c++, mpi, cuda, basics of networking or distributed programming) == 🦄
What extent of cuda knowledge is helpful? I know the basics but haven’t done anything hands-on professionally
Didn’t know data scientist ever needs to use c++ and networking/distributed programming. Always thought these are for swe
This might be due to the fact that DS is not a well defined position. In many companies the business intelligence is done under DS which effectively only needs SQL and tableau or a similar tool. I appreciate one needs to spend a lot of time to master these two tools, but in order to pass the interview or do the job it’s much easier than learning all aspects of ai engineer . DS roles are much less well defined than an AI Swe
Yeap. There’s some better definition happening. For example at Google we recently split data science into data science product and data science research. DSR is now benchmarked at 50% DS and 50% research scientist pay.
I second this. "Data Scientist" is too poorly defined and often includes data analysts and business intelligence who don't really work on production systems.
because all the bio and psych majors flooded in when they realized being a labcoat is a shit job and gimped you with imposter syndrome
Cos most DS roles can be replaced by import sklearn
Sklearn is garbage when you're actually trying to make inferences. It was definitely designed by engineers rather than statisticians.
Inference is not the same as prediction. Sklearn does prediction and it does it well.
As others have said, maybe it's because DS means different things from one place to the other. BI and Data Analysts are typically not well paid but Research Scientists (and all things ML adjacent as you said) are. Most companies that I've been at who kept the actual "data science" title in 2023 have higher pay bands for DS compared to swe
Most DS positions are glorified data analyst roles. If you never learnt C++ in your life, you don't know how many tools actually. Most people learn R or Python in a week and think they can code, but that's not enough.
You need to know C++ to do data science?
You do not need to know C++ to do real AI/Ml lol
Tech Industry
11h
2174
Tech companies to avoid as a white guy?
Tech Industry
11h
1450
Lack of diversity in engineering division at X
Tech Industry
16h
1050
Tiktok offer
Tech Industry
7h
467
Why do families communicate so poorly
India
6h
463
People who like Modi, are you okay with your country’s leader using words like “infiltrators” for a community?
Is intel hiring for DS
I think there are a couple of jobs for it. Most people would not recommend coming here though.